During web surfing happened over this picture – Deliberately not on a separate line in order to make the forum software not infringe on copyrights.
The picture shows construction of either a MTB single trail or a hiking path, I cannot tell for a certain. It shows a “Bagger” (German term) and a compacting device that I do not know what it is called, neither in German nor in English.
What tag will the resulting surface receive?
I’m “winging it” with this answer (I’m not a professional road grader / builder) but as I see that there is “grading” equipment (medium/heavy-duty-equipment to “compact” the roadway to provide a smoother surface and/or easier gradient, especially as the visible switchback does exactly this), it could be said to be “compacted” (in my opinion). However, a MUCH closer view of the actual surface of the road would be helpful to see to reach a final conclusion, as a “compacted” roadway (could be sand, gravel, dirt…) has a “compacted” look to it (having been actually compacted).
I realize this only might get us closer rather than conclusively defining “compacted.”
Grading and compacting are distinct operations. Grading shapes the profile; compacting reduces air voids under pressure. You can grade without compacting the result stays loose. What determines the tag is the finished surface condition, not the machinery present.
The same as with at least one other surface threads with a single photo not proving enough info you made: not enough info, impossible to say.
It is not possible to guess final surface of path based on in-construction photo taken from far away.
If you plan to actually map this surface you need to survey this path after construction ends, and not from other side of valley.
to elaborate:
it is under construction, this segment may end with surface=asphalt or be covered with soil again and is serving now only as access road - to take two unlikely extremes
we have no closeup
AFAIK fair use would apply here anyway
especially as machinery may be not working on this specific section and only being moved to other location
Thank you! I stand corrected.
Isn’t the OSM definition that it consists of dirt and pebles that have been compacted together?
If they don’t add any materials, I think it would be surface=dirt, but with a very nice smoothness.
The wiki definition of surface=compacted is ambiguous: it starts with a specific engineered surface (water-bound macadam) but then broadens to something often called “gravel.” In practice, it’s become a catch-all for “good unpaved,” making it unreliable for material-based queries or routing.
If no clear added material is present, map the best identifiable ground type (e.g. dirt, clay, gravel, pebblestone) and use smoothness=* to describe ride quality.
To one person here a scene as in the photo seems to be like from another planet. In retrospect, that may well haven been a reason for me to share that photo, because it shows something that openstreetmap contributors rarely happen over.
From talking with people who construct such trails/paths: Whenever possible, they use material sourced from the site. They have the machinery to dig that out and the machinery to carry it to the place where it is of use. This is not always possible, where the material that can be locally sourced cannot get compacted. Then they have to import material that allows them to create a decent surface. There is no way for somebody not being there during construction to know how they decided. An expert in track development perhaps could dig a hole into the surrounding ground (in my area mostly humus) to look underneath and give an estimate. How many contributors to openstreetmap willing and able to do that?
From looking at what is tagged in the wild: If surface at all, then mostly “ground”. I guess in the US it would be “dirt” instead? Some may tag “earth” too. Regarding my question: @julcnx coming closest to an answer.
Not a track builder myself, I’d say Humus can be compacted to a become hard as a bone, The resulting surface though not durably so, e.g. after heavy rains or during snow-melt season.
they use material sources from the site
What I mean is, that “compacted” implies particles of different sizes - such as dirt and pebbles.
It shouldn’t matter if this mixture occurred naturally or was placed with intention, nor where the materials came from.
The surface described in this post should perhaps be given a new tag like “surface=compressed_dirt” and a tracktype.
In plain English, “compacted” means compressed into a denser form, it describes what was done to a material, not what the material is. Clay can be compacted. Dirt can be compacted. Gravel can be compacted. So arguing that origin shouldn’t matter only makes things worse:
This repeats the same problem: encoding preparation into the material tag. If the material is dirt, the tag is surface=dirt. Whether it was compressed, graded, or left natural is a separate dimension and belongs in a separate key, not baked into surface=.
Otherwise we open the door to an endless proliferation of compound tags: surface=loose_gravel, surface=graded_dirt, surface=packed_sand…
Sure, there are plenty of surfaces. But will that help consumers when there are as many tags in the data? Below three photos from a site where new trails are laid:
I certainly won’t fail a mapper that maps surface=earth there, or dirt or ground. It does not tell a consumer much about what is there, but it is not wrong neither. Just noticed, I did that myself on several occasions.
Not sure how I’d have tagged such when on a forestry track. Plate compactors are also used when building those to get the crowning in one go.
Compacting dirt will likely change smoothness but not surface
And dirt with earth are synonyms
Did not take long to find a trackroad surfaced with the same material:
It is of course tagged surface=compacted. There are 20 ton lorries driving several times a workday all year round there to a waste recycling. So certainly has a solid base layer underneath.
Looks like surface=dirt + smoothness=intermediate + tracktype=grade2 to me !
for me (based on photo and description) looks like surface=compacted
The surface is mostly fine brown material with only scattered small stones, possibly imported. Without them, this reads as compacted earth.
It looks like compacted brown gravel to me, with admixture of brown earth.
And matches some nearby ones close to me.
Hard to verify, that surface could equally resemble a rolled laterite track near me. The wiki doesn’t distinguish engineered aggregate from graded native soil, which I’ve flagged here: https://wiki.openstreetmap.org/wiki/Talk:Tag:surface%3Dcompacted#Definition_mixes_material_and_metho
Perhaps I should say, the pile of earth here does not contain organic material, it is all minerals, so I was told. When explained whether it compacts well I was shown that one can form rolls from it and they keep shape, at least when wet a bit and pressed hard, kind of a snowball. Likely there is some clay in too, these mixtures must start with the smallest diameter particles. I heard it called “Wegsand” (literally way sand) too, certainly not in line what sand means in Wikipedia.
Most surfaces go by the name of a material. There is no material called “compacted”, compacted is rather like paved. Certainly, gravel, as understood by OSM, cannot be compacted. Mixing with sand of no use neither. While dirt sometimes can have all the properties needed for a durable compaction 
Personally, I’d only call that dirt when somebody litters that on the floor of my flat. On site it does not look displaced, what I consider the essence of “dirt”. Earth maybe is too broad in scope, can mean humus too, the organic stuff, the top soil that is put into flowering pots? Ground (another synonym?) maybe right when the material was not imported but dug out from beneath the humus in the surrounding, but how can mapper Joe or Jane tell that? If it was a hiking path the builders certainly would throw a bit of fine gravel on top of it (many mappers I guess still think that this is the definition of that surface tag, and many consumers still think so too, not without merit, I’d say.)
PS: So tracktype to the rescue? The grades (in my opinion, and I think that I am not alone in that), that the feature or scar in the landscape as @Ben coined it, was purposely made to be usable by as many as deemed to be the users and in a frequency they deemed to be rightfully using it. In my opinion this latest photo not a track at all but a service (because of most prominent use), from what I learned in other topics, tracktype also consumed for those. Nice to know whether information gets lost or not when changing primary tag.
