Talk:Tag:surface=gravel

Other potentially useful images:

• 
  Relatively loose relatively large gravel - but not too much.
• 
  Gravel sunken and compacted with dirt, forming relatively nice surface
• 
  Small loose gravel
• 
  Mix of sunken an loose gravel

Mateusz Konieczny (talk) 07:30, 13 March 2021 (UTC)

More examples

Gravel may often be used as a synonym of macadam according to this old discussion on related surface types.

• 
  Pure gravel
• 
  Loose and abundant gravel road in Czechia
• 
  Loose and abundant gravel road in Estonia
• 
  Somewhat loose gravel road in Iceland
• 
  Close up, for comparison with pebblestone
• 
  Gravel pedestrian paths in a courtyard in France
• 
  Gravel street in Brazil
• 
  Gravel-covered dirt road in Brazil

Fernando Trebien (talk) 18:40, 21 March 2021 (UTC)

Please make this more concise

I have seen "surface=gravel" on paths like this http://www.mountainbikers-paradise.com/tracks/lafatscher/fk/lafatscher29.jpg (the tour guide uses OSM http://www.mountainbikers-paradise.com/track/innsbruck/lafatscher) and where paths cross a scree, cant remember how many of those I let mistakenly drop. --Hungerburg (talk) 22:49, 15 April 2021 (UTC)

Note that short version of surface definition is at Key:surface, though if some parts are not needed and can be removed without losing something useful - feel free to make an edit Mateusz Konieczny (talk) 05:49, 16 April 2021 (UTC)

Perhaps, I worded my statement badly; I let drop the tag from the data; Eg. on Lafatscher Joch, otherwise all of the apex there would be surface=gravel now, while actually there is some gravel/scree on the way, but there are also rocks 10cm and more and a few 30-50cm steps of bare rock too. Would you agree, that gravel is a kind of cover for compacted? To me, all the pictures above seem to point at that. --Hungerburg (talk) 14:27, 16 April 2021 (UTC)

"Would you agree, that gravel is a kind of cover for compacted?" - sometimes amount of loose gravel is very large, sometimes there is only a bit of gravel and earth otherwise Mateusz Konieczny (talk) 18:25, 18 April 2021 (UTC)

I added a picture to the gallery that may or may not show a mountain path, I can see it! I'd never map there a surface=gravel, but in the data there are samples, and indeed, it is loose, as the definition does/did state.

Thinking about, how to disqualify the mountain path example - only this comes to mind: the gravel was not put there by humans. It is unpaved gravel, like surface=earth, while the other pictures might show paved gravel. --Hungerburg (talk) 19:10, 18 April 2021 (UTC)

I would say that such naturally occurring gravel can actually qualify for surface=grave (it is hopelessly broad anyway...) Mateusz Konieczny (talk) 20:09, 18 April 2021 (UTC)

Rocks of 10–50 cm fall above the gravel range by standard geotechnical scales: ISO 14688 and the Wentworth scale both place the gravel/cobble boundary at 63–64 mm. A path with exposed rock steps and 30–50 cm rock outcrops is surface=rock, not gravel. Proposing to document this upper boundary explicitly, see #Upper size limit below. julcnx (talk) 13:54, 4 May 2026 (UTC)

Upper size limit

The page acknowledges a "very wide meaning range" but sets no upper boundary. Standard geotechnical scales (ISO 14688, Wentworth) place the gravel/cobble boundary at 63–64 mm. A surface where loose fragments larger than this dominate is better described by surface=rock.

Proposed addition: state explicitly that the dominant-particle range for surface=gravel is approximately 2–63 mm, and that surface=fine_gravel covers the lower end (approximately 2–6 mm).

See: https://en.wikipedia.org/wiki/Gravel#Definition_and_properties

— julcnx (talk) 13:45, 4 May 2026 (UTC)

I agree with that and have been using that definition in my mapping since 2014, although larger rocks/ballast are very rare in my mapping region. --Fernando Trebien (talk) 14:12, 4 May 2026 (UTC)

After further research: setting a hard 63 mm cap on surface=gravel would invalidate widespread existing usage: central european forest tracks and rural roads surfaced with large angular crushed rock (including material exceeding 63 mm) are routinely and reasonably tagged surface=gravel, and the German Atterberg engineering scale classifies all such material as "Kies" (gravel) up to 200 mm. Imposing a strict geotechnical boundary would break backward compatibility without community consensus.

Revised proposal: keep surface=gravel as a broad catch-all with no enforced upper size limit, consistent with its current documented role. Document the 63 mm geotechnical boundary (ISO 14688 / Wentworth) as a reference scale, not a tag boundary. Introduce surface=ballast as a precision opt-in for surfaces where dominant stones are 63 mm or above (Wentworth cobble lower boundary, consistent with ISO 14688-1 Co ≥ 63 mm) and cycling is near-impassable. julcnx (talk) 11:36, 9 May 2026 (UTC)

I believe cyclists would generally prefer to avoid riding on both small-particle gravel (prone to sinking due to the thin tires) and on ballast (which throws the cyclist off balance). Cycling is generally easy on compacted, acceptable on dirt when dry, and difficult on sand (prone to sinking even for cars). --Fernando Trebien (talk) 18:01, 11 May 2026 (UTC)

The difficulty mechanisms are different. On ordinary gravel the obstacle is rolling resistance and sinking: slow and effortful but controllable. On ballast-size stones (≥63 mm) the fragments are large enough to suddenly deflect a wheel sideways under load, causing falls that are independent of tyre width or rider technique. That is the routing signal the ballast proposal targets: not merely "slow" but "unpredictably hazardous." The two surface types are difficult for cyclists for different reasons and to a different degree. julcnx (talk) 03:48, 12 May 2026 (UTC)

Dominant character

The page does not address what happens when a surface has mixed-size material. The surface=compacted counterexample already encodes one direction: coarse gravel dominating = surface=gravel by community decision, even if some fines remain. The same principle applies in reverse: a bound and compacted matrix with a few scattered coarse or cobble-sized stones on top remains surface=compacted. Those pieces are incidental and do not override the character of the surface a wheel rides on.

Proposed addition: a short paragraph stating that the tag reflects the dominant character of the top surface, with both directions made explicit, cross-referencing the counterexample on surface=compacted.

— julcnx (talk) 13:45, 4 May 2026 (UTC)

Makes sense to me. --Fernando Trebien (talk) 14:13, 4 May 2026 (UTC)

Document natural vs. imported/crushed gravel as distinct sub-types

The page describes surface=gravel as having a "very wide meaning range" but does not state whether it applies to naturally deposited material, crushed/imported material, or both. In practice both are used: bank run gravel (naturally deposited, also called "river run" on Wikipedia), excavated from a riverbank or pit, and crusher run (Crushed stone on Wikipedia), an engineered product. These differ in shape — natural gravel is rounded by river transport, crushed stone is angular — which affects how the surface binds and behaves under load. Documenting this distinction would also correct a downstream error on surface=pebblestone, which currently defines surface=gravel as "produced by crushing... has sharp edges," excluding bank run entirely. julcnx (talk) 07:32, 9 May 2026 (UTC)