Talk:Tag:natural=mountain range

Why not areas?

I think that for MANY mountain ranges, mapping as an area makes the most sense. Look at the Olympic Mountains in Washington state, for example: there is no single line that could be drawn that would make any sense, map-wise. And mapping as a single node doesn't fit for moutains that extend across entire continents. (Where does one place the "node" for the Andes?)

It's true that with areas, there is some question of how many foothills to include, but I think precision is less important than getting the ranges on the map in a way that map renderers can do something useful with them.

--Adam Schneider (talk) 23:47, 26 May 2019 (UTC)

While it might be useful to be able to map mountain ranges as areas, the geometry often lacks verifiability. As you said, it's hard to decide how far out into the foothills the Olympic Mountains extend. In California, the transcontinental railroad managed to make a bunch of extra money by redefining the Sierra Nevada mountains as starting right outside of Sacramento, at the first low foothills - the government couldn't really prove that these were not part of the mountain range.

If you look at websites that catalogue mountain ranges - like peakbagger.com - you will see that they have the area of a "range" extend all the way down to the sea or a river on each side, which includes large areas of flat valley floor in the "mountain range" area. These are not "real and current" features that we can map in OSM. Their area map for the Olympics includes the entire peninsula, down to Aberdeen. This might be a way to map mountain ranges as an area, but it's probably not what you were thinking of when you imagine "The Olympics Mountains".

While the foothills and lowlands are debatable, the highest peaks and ridges of a mountain range are certainly part of the range. This is clearest with simple mountain ranges that have one main crest, like the Sierra Nevada. These can be reasonably mapped by a line that follows all the ridges and peaks that form the crest.

For example, I mapped the Siskiyou Mountains in Northern California (and Southern Oregon) as 2 linear ways along the highest crest, but it would be quite hard to decide where they end on the north-western side if you try to map an area. Compare to peakbagger's area, which reaches the Klamath and Rogue rivers and the Pacific Ocean as borders to the mountain range, even though this includes large areas of low, flat land: https://www.peakbagger.com/range.aspx?rid=12231

To fully map the Olympics, I would stick with a node for the mountain range, at the top of one of the main ridges at the center of the area, and then I would map any smaller, named mountain ranges that make up the Olympics, and individual natural=ridge (named or not) that connect all the high mountains. This would provide enough information for a computer to estimate the size of the mountain range

For the Andes, there area really several different mountain ranges with that name, if you look at a topo map, sometimes separated by valleys or plains. But the individual parts of the Andes could be mapped as mountain ranges, for example the Cordillera Occidental, Cordillera Central and Cordillera Oriental in Colombia might each be mapped, though it would be even better to map the smaller, local ranges that make up each of these long Cordilleras.

An individual map renderer who wants to show a label for the Olympics will be making a map of a large part or all of Washington State. It's easy enough to adjust the size of the label by hand. The one problem is for maps that want to render the whole world by computer, without using any local knowledge about the size of the mountain range. In this case, it would be necessary to using a DEM to estimate the size of the mountains, say by finding the area above 1000 meters around node. Opentopomap.org already uses the DEM data to orient the icons for saddles and passes, and to calculate the isolation of peaks for labeling, so this would be feasible. If all the ridges and peaks in the mountain range were mapped, it would also be possible to follow these out from the node for the mountain range and get a good estimate of it's importance, to set the initial zoom level or label size.

But we should not base mapping and tagging decisions on what features will be rendered, rather we should try to map what is really there, and let renders figure out a good way to represent this. -- Jeisenbe (talk) 03:06, 27 May 2019‎ Jeisenbe

Only the use of an area does make sens to me to have an aproximate definition of the region. Maby a node, but that's almost too unspecific. As a line? I think many people mismatch "range" with "arete" or "ridge". --EinKonstanzer (talk) 14:19, 28 September 2021 (UTC)

I think areas are just fine even if you don't know the exact borders. For the way (line) you also don't know the exact center of the mountain range and where exactly does it start and end. Also if you have mountain range of a round shape then single line can barely show its shape. And the node will tell you nothing about its size. So for me the area is the right way to go. There are also many forests (natural=wood) where trees are just sparser and sparser. Also there you can't tell where exactly does the landcover end and it is ok to subjectively decide. Anyway there should be some fuzzy key proposed. --*Martin* (talk) 15:24, 14 September 2023 (UTC)

Link to place=region

I don't think there should be a link to place=region in "see also." The tag region:type=mountain_area has never been documented, from what I can see, and has only been used in a few places (see https://taginfo.openstreetmap.org/keys/region%3Atype#map). --Jeisenbe (talk) 08:04, 21 October 2019 (UTC)

I have added the page DE:Tag:place=region. It is not a 1:1 translation of the English. Maybe an url translation helps.

I think 1. Tag:natural=mountain range means the same as 2. region:type=* + region:type=mountain_area. The version 2. comes at least 376 times that is not so little. To map this attribute on a point is not good, because then the extent is not recognizable. Maybe we can improve the definition on the wiki page even more?--geozeisig (talk) 15:31, 21 October 2019 (UTC)

If there is a link on the English page, there should be an English-language description. The German page doesn't really have a definition or description of region:type either, it only just mentions that region:type=mountain_area is used for a "Bergregion" which I think means "mountain region", so that's not really helpful. But based on usage, I don't think this is at all the same as natural=mountain_range, which is a linear feature made up of ridges and peaks. The place=region areas in the alps seem to usually include several different smaller natural=mountain_range features, and also include many valleys, lakes, towns and even cities. These areas do not have verifiable boundaries, so it's not a good idea to map them as area in OpenStreetMap. --Jeisenbe (talk) 00:12, 22 October 2019 (UTC)

More then 2.500 times used now

There are not that much mountain ranges in the world! A bunch of hills is not a mountain range. A much more sharper definition is needed! --EinKonstanzer (talk) 14:10, 28 September 2021 (UTC)

But eg Mountain range on Wikipedia lede starts with "A mountain range or hill range is a series of mountains or hills ranged in a line and connected by high ground. A mountain system or mountain belt is a group of mountain ranges with similarity in form, structure, and alignment that have arisen from the same cause"? The article has a separate section for Mountain range#Major ranges on Wikipedia. The definition of a "mountain" or "hill" itself is not "sharp" either. Coincidentally the Tag:natural=hill suggested a natural=hills for natural=hill.

If you mean "a bunch of" natural=peak and natural=ridge shouldn't be a natural=mountain_range, then we need a natural=mountain to group them in between.

---- Kovposch (talk) 09:54, 29 September 2021 (UTC)

Relation?

I think it could be useful to have mountain ranges be relations in a similar way that waterways can be. Sometimes the ridges of named mountain ranges are separated by rivers and watersheds, and it could be useful to have these two separated lines be connected by a relation; perhaps prominent peaks could be tagged as part of the relation as well. Any thoughts? I'm not really sure about how this could be done, or how the relation could be created, so I'd be interested to see what everyone has to say. --AragonChristopherR17z (talk) 07:47, 19 December 2022 (UTC)

I think creating relations for mountain ranges makes a lot of sense. That would allow us to handle a mountain range in the shape of a "Y", for example. Instead of just being one line of peaks, it could include branches.

--Adam Schneider (talk) 02:57, 17 January 2023 (UTC)

I also agree with the relations. See also my comment to areas. --*Martin* (talk) 15:25, 14 September 2023 (UTC)