Trail_visibility explained

That’s not how I interpret that phrase. For me, “well cleared” means “free from obstacles” (that could twist your ankle if you’re not wearing hiking shoes).

I think according to the context, well cleared doesn’t guarantee there aren’t any obstacles that can twist your ankle, it means vegetation has been removed, as well as most obstacles (compared to “normal”), but in alpine regions the typical (normal) situation is the requirement for hiking shoes even after it is “well cleared”

Whether the path is free from obstacles because some human removed them or because nature never put them there is irrelevant, I think.

“cleared” implies they have been there and have been removed, otherwise it would be “clear”

1 Like

The majority of single track trails in the Sierra Nevada are neither free of anything which you can trip over, nor proper T2 where you are moving over a constant uneven surface. Just because there’s some obstacles in the trail doesn’t mean that it’s hard to see where it goes or that you need boots - many times you can walk around obstacles on a trail. It’s also pretty common for people to hike off trail in rugged terrain with low top approach shoes. My partner will carry a heavy pack on T2-T4 terrain with minimalist trail runners.

I also read well cleared as that vegetation will be cut back and branches removed from the trail, not that every single rock and root has been taken out of the path. In order to be visible it just needs to be easily distinguishable from the ground around it - a trail can have a lot of roots on a dirt surface and still be clearly distinct from thick vegetation on either side.

None of the photos on Key:trail_visibility - OpenStreetMap Wiki show trails completely free of obstacles.

This sounds like my proposed hiking_technique. I’ve been thinking about the obstacle issue I outlined above over the summer, and someone else in the thread made a point about it as well.

I took a brief stab at this earlier by looking at a variety of international bodies that create trails, feedback from other threads, and my own experiences.

It’s a very rough draft and needs some editing and condensing, but I think it does a good job of establishing a truly “excellent” excellent and not getting bogged down in a lot of confusing overlapping lower ends of the scale.

In regards to the earlier comment on exposure, I’d be fine with just dropping exposure from difficulty but not creating a new key and just using hazard=* nodes though there’s some things to sort out there.

@Hungerburg many months ago I mentioned some photos of a trail that fades - here two sides of the same trail at a meadow where it loses definition (not enough foot traffic to overcome growth)

The rest of the trail is always solidly visible, so I wouldn’t downgrade the entire way, but it also seems annoying to split into a new way with trail_visibility=no for ~150m of easily navigable open terrain.

If I was to follow this trail guiding myself with a map, I’d be glad to see the trail going fainted on the map on this section, so I would trust I’ll find it again across the meadow.


There’s a LNT take that there shouldn’t be a path shown there, as concentrated use could lead to rutting in the meadow which would impact drainage, could kill off areas of grass, etc. It’s unlikely to happen in this case given where it is and the unlikeliness of it becoming very popular but it’s definitely not a case where you need concentrated use to reduce braiding.

For routing purposes a way/segment with trail_visbility=no makes sense.

Well, that is the single most reason, why the tag trail_visibility=no exists at all: To map easy hiking sections over pathless terrain. (From 2012? ML discussions on why that key got split from sac_scale.)

If there was a gap there instead, you would just as well find the continuation. But you could not let routing software guide you there.

Yeah that’s the only legitimate use of it I can think of, otherwise it’s a pathless route masquerading as a path.

Ideally a map renderer wouldn’t display it but routing would work - it’d be better if people can see either side of where the trail exists and make their own way across it to spread out damage to the alpine meadow since the trail has regenerated. Since there’s a path there now anyways and there won’t be many people on it I’m probably just overthinking it. :slight_smile:

Alpine meadows are a thing of the past. Alpine pastures still exist. Sounds like a play on words? A marketing person called them something, lots of other places look with envy at.

It’s more : following the dots, at least on the map, is a small pat on the shoulder.

This discussion is getting a bit confusing because it’s both about trail_visibility and sac_scale. One thing they have in common is that their first level is based on the German phrase “gut gebahnt” which is translated to English as “well cleared”.
On our hike yesterday we came across this path. Can I assume we all agree that is a good example of “gut gebahnt”?

I would certainly tag it with trail_visibility=excellent because it’s dead obvious where the path is. I think this is because the path has been enhanced by human hand: it has been cut into the slope, maybe by a bulldozer. I think this is what distinguishes a trail_visibility=excellent path from a trail_visibility=good path. A trail_visibility=good path is also easy to see and follow, but it has not been shaped by human hand. It is like an informal path, i.e. it “has not been established on purpose” (from ). A trail_visibility=excellent path, however, has been enhanced by human hand, because it has been bulldozed, has been paved, gravel added, secured with a fence, etc.

Similarly, I would tag this path with sac_scale=hiking because there is nothing that makes it technically demanding. Reading the wiki, the most notable difference I find between sac_scale=hiking and sac_scale=mountain_hiking is the recommendation of hiking shoes and some sure footedness for the second. I think the thing that makes this necessary for sac_scale=mountain_hiking is the presence of “obstacles” (still looking for a better word, maybe “roughnesses”?). The “gut gebahnt” of sac_scale=hiking can then be interpreted as their absence, i.e. it is “clear” of them (by human hand or naturally doesn’t matter). Of course no hiking trail is completely clear of them (there are some pine cones and a tree root on the photo) so there will be many cases where a mapper will be in doubt if their number justifies tagging it as sac_scale=mountain_hiking or it is a case of the “bad end” of sac_scale=hiking. If there are only a few and they can easily be avoided, then it can still be tagged sac_scale=hiking. I also noticed that on this path, we were walking much faster than on the preceding sac_scale=mountain_hiking path, so maybe the average speed a hiker can make on a path could be an additional criterion to decide between sac_scale=hiking/mountain_hiking (on a sac_scale=mountain_hiking path, you need to slow down).

Eh, the wiki states that good “sometimes has to be searched for.” Honestly both of the example photos for good look more like intermediate to me (a very faint path through grass, then another which doesn’t exist for 50 meters), but we’ve complained about this before. :stuck_out_tongue:

I think based on the current key that the path shown there would fall into excellent visibility.

Tagging it as T2 doesn’t seem appropriate, as someone doesn’t need balance & good footwear to step on leaves, but SAC is full of weird little requirements. T2 is required to be steep, T3 is required to have exposure. Neither of these requirements are reflected in either sample photo on the OSM wiki. The T3 photo also does a poor job of showing “Use of hands for balance potentially needed” as the only time you’d really use your hands on that terrain would be if you fell as the majority of the obstacles are below thigh high.

PS - If we take into account how visible the path would be in bad weather (which multiple countries trail classifications do, along with the first example here it’d be less clearcut but the vast difference slope would likely keep it easily followable. The following probably has little chance of being adopted as it doesn’t map well to the existing system, but it feels like it makes more sense in terms of how the average person thinks of path visibility and ensures that excellent is actually excellent not just good. RFC: hiking_visibility key 🍿 paired with hiking_technique and hiking_exposure/fall_risk this rounds out the hiking path trinity

PPS - @Hungerburg I seem to recall somewhere you asking for examples of T2 and T3 terrain outside of mountains, but can’t find where it is. One easy example would be coastal areas, which often have boulders. I’ve posted lots of photos from the southwest which contain T2-T5 terrain despite not being mountainous (the sandstone ridges are only a few hundred meters high, but tend to be steep). I have relatives that live on the east coast - there’s an informal trail up the side of a hill that has a small rock face that is T4 (the hill is ~330m high, the slope starts steepening after 265m and a lake nearby is at 170m).

The ‘well cleared’ language has to be leftover from when SAC included trail visibility. These trails, while not having obstacles have zero construction as far as I can tell and are just repeated tread.

Technique wise they’re dead simple. T1. Visibility wise they definitely aren’t excellent.

FWIW in terms of the photo now at the bottom of the sac_scale page I’d put it at T2. It’s flat and not in a mountain, but it seems like people have to hike over the roots and rocks vs around them which would make it clearly T2 in the part that actually counts - needing good balance and traction to move over terrain.

Actually, it speaks of steady ascent of the trail, and that’s a bit the opposite: Ascent spread evenly along the route. The picture in the Wiki is really telling, T2 is about walking skill. Our local alpine club offers mountain hiking courses and they are exactly about paths as the one below:

I put that there and the wording from the definitions above besides it. I too think, they match very well. Look where there is no moss on the rocks. Curiously, in a poll in the German forum, 16 out of 31 votes went for hiking. I guess, because the scene is not set in the mountains. I opened that poll, because in another topic a picture of a similar scene was posted with the question, how to tag that, and putting on sac_scale=hiking got recommended.

Meanwhile I prefer well paved as a translation. The German Bahn is something on the ground, that makes it stand out from the surroundings. If this is trail_visibility instead, then there is nothing left to recognize T1 except, exposed areas well secured, so, can be anywhere?

I’d say that having found the continuation from looking at the terrain, orienting and reading the map is a lot more satisfactory than having my position on a device match the path drawn there. I hope you do not do this in alpine terrain.

To repeat my thesis from the beginning, you can map a gappy highway=path that passes pathless terrain, then close the lids of your eyes until everything starts to blur, then you get the OSM trail_visibility value - in this case likely good, as we all know.

Tagging like that certainly would make openstreetmap data much more usable for lots of consumers. As of now, it is only useful for wilderness lovers, and of not much use for ordinary people just out for recreation.

I agree that it should just be walking skill, but that picture is of a lakeshore. Lakeshores tend not have a steady ascent due to water filling in terrain.

A steady ascent could be an evenly spread 1% angle or an evenly spread 45% angle. It’s a weird distinction to make, I guess they just don’t want someone getting worried about a brief 10m rise over otherwise flat terrain?

update: According to the OSM Wiki, T2 is “Terrain is steep in places” not “steady ascent”.

Yeah SAC scale naming is dumb outside of the context of the SAC.

This was in the desert not the mountains, so it’s T1.

Maybe we need a demanding_desert_hiking value? :stuck_out_tongue:

This is how I think of it:

T1 is simple walking on a trail with or without obstacles here and there that are mostly avoidable. This can be anywhere.

T2 is surefooted walking on top of obstacles, or up a steep enough angled slope that it’s sort of it’s own obstacle that you need some experience to pick your line, sidestep, crossover, etc.

Whether or not it stands out from surroundings or not doesn’t impact technique necessary just visibility. It’s not like the trail of trampled grass is harder to hike on than if it was compacted dirt with grass around it.

I think insisting that T1 can’t have any exposure or fall risk, and any that exists should be roped off / secured gets you into weird edge cases. I also recognize this is how SAC works.

So what is this?

It’s walking on a nearly flat surface, so it’s T1.

While not knife edge exposure by any means it definitely isn’t secured or roped off, so it’s at least T3.

But… it’s not in the mountains (lol), so it’s T1.

Routefinding is fun!

Not everyone is comfortable routefinding.

Yeah, it’s meant for members of an Alpine Club after all.

Seems like we’re close to consensus here!

I don’t think it’s “required”: I don’t think the wiki description should be taken literally, instead the best match should be found to what’s on the ground. I think one of the problems in the OSM comunity that leads to long discussions is that many of us are also IT experts who are used to think “digitally” and like to write “code” in the wiki that a computer can understand so it always knows what to do. However, the world we are trying to map is an “analogue” world with an endless variety of circumstances, and we’re mapping it for map users that are not usually used to think “digitally” either. So we should be more free in our interpretation of the wiki, and apply it with some common sense. The sac_scale key is meant to describe hiking difficulty, and the wiki describes what paths of different degrees of difficulty usually look like. But there will always be plenty of exceptions: paths that are tricky to walk on but are not in the mountains, paths that are free from obstacles but will make you dizzy with vertigo so still quite difficult to walk on if you suffer from it, etc. So if we try to map a path that’s strewn with 1 m high boulders we have to climb over, but doesn’t go up, we should not go into “error mode” but just tag in T3 because we need our hands to get over those boulders even though you could drop max. 1 meter so there is hardly anything you could call “exposure”. What we should strive for is a wiki that describes how to tag paths so that we get a ranking of paths according to their hiking difficulty that most map users would be happy with. What we want is that our customers (hikers) say “That T3 path we just did is really harder than the T2 path before and good we didn’t go to that T4 path: thanks OSM job well done!”

1 Like

I agree with you, but there’s at the very least a significant minority of people that view sac_scale very strictly.

Having all of the secondary aspects (areas must be secured, or well cleared, or at a certain slope angle, or in the mountains) just makes it more ambiguous and harder to tag in a consistent manner. If there’s a trail which is clearly T1 difficulty, tag it that way and then put a hazard=steep_slope or something on the side if it isn’t fenced off vs bumping it to T2.

Does it though? There’s one sample photo per level and they’re not always instructive, the T3 photo is pretty marginal for using hands for balance. They certainly don’t do a good example of showing what they typically look like in the desert southwest of the US. It’d be better to at least include a few different examples of different archtypes of terrain at each level.

I put up some examples of terrain that can be hard to classify (steep slabs, or steep loose terrain) - if there’s some consensus around that examples like that could be included etc.

The following photo to me seems to indicate T2 terrain pretty clearly - it honestly looks slightly more challenging than the sample photo for T2 where someone is walking along a lakeshore with their hands in their pockets. A slim majority of people thought it was T1. This is presumably largely because of the secondary aspects in SAC.

I agree! It would be both fun and helpful to make a gallery for sac_scale, and maybe another one for trail_visibility, similar to this one, with examples from different regions around the world, with different terrain and path surfaces…

This could help people decide which tag to use, and it might even pave the way (hehe) for quests on these tags to be included in StreetComplete and similar apps.

1 Like

We are drifting off-topic, still the conversation is entertaining:

I did start that here - right side of table deliberately not grading, only mentioning terms from the definition and presenting what they may look like on site.

While SAC/CAS® is quite clear in its hand-outs, the highest grade always wins, OSM mappers sometimes choose a lower grade, due to such secondary aspects. The poll in the German topic – from replies to my questions on why – shows another point of failure: Some people regard sac_scale=hiking as a means to tell, this is for recreation, an equivalent of hiking=yes perhaps.

You are too optimistic! While I agree with the goal, making openstreetmap data useful to the general public :slight_smile:

I understand your desire to map each axis that has to be considered separately, as installed in several topics; This should prevent failures as above. No idea though, if mappers will do that extra work or consumers will follow suit - they’d have to create a lot of UI to make that usable.

The only thing that needs to be disambiguated is exposure - IMO everything else is already covered or is irrelevant (slope, biome, visibility, etc).

I’m fine with expanding the hazard key into hiking examples (which should be done anyways) to cover that instead of creating a new key. There’s been some good discussion in the exposure thread I created a few months ago. :slight_smile: