OpenStreetMap

Current natural=water scheme inconsistency

Posted by BushmanK on 8 April 2016 in English. Last updated on 6 February 2017.

It looks like the current scheme for tagging “water-containing features” has certain significant inconsistency. This scheme originates from this proposal made by @Zverik back in 2011. I completely support the idea of having some uniform scheme for more or less natural water bodies, but this proposal also has an effect of deprecation of several tags, related to artificial, man-made or even industrial objects, where water is in use.

This issue has been addressed in original proposal discussion by @G0ldfish, but @Zverik didn’t answer his question at all.

After approval of this proposal, the whole landuse=reservoir scheme has been labeled as deprecated, because supposedly, we can now use natural=water water=reservoir to tag artificial water bodies instead of using landuse=reservoir. But it’s only partially true. Indeed, water storage reservoirs are often not that different from natural water bodies. But there still are objects such as cooling basins, technical water reservoirs, wastewater reservoirs etc. Important tag reservoir_type= is deprecated as part of the whole old scheme.

Please, keep in mind, that when we have to tag a cooling basin of an industrial facility, water itself is just a part of this object. Ideally, proper tag (something like former landuse=reservoir reservoir_type=cooling) for it should indicate that there could be a concrete basin with sprinklers and other equipment. It can even use some sort of cooling liquid, which is most likely water-based, but it is absolutely not necessarily a water.

Same situation with sewage and wastewater reservoirs. It’s quite hard to call its content “a water” - water usually is a product of long filtering process, and aeration in sewage reservoirs is just a part of it. So, initially, concentrated sewage liquid can not be called “water” by any standards.

So, here we have the case, where a proposal has not been discussed and tested in every detail, which leads to the situation, where artificial objects were forcibly brought into the same class with more or less natural objects. Another result of it is that now we have less detailed options for tagging artificial objects. Fortunately, depreciation doesn’t mean automatic re-tagging or prohibition to use this scheme (as it is mentioned in proposal text), so it is still possible to use reservoir_type=, for example.

Personally, I think there is certain sense in reviving man_made=reservoir tag for obviously artificial objects such as sewage treatment tanks, cooling basins and so on, where the whole vessel is formed by concrete or steel walls, since these objects have nothing in common with natural water bodies at all, being artificial and not containing a water.

Discussion

Comment from Warin61 on 8 April 2016 at 23:04

One problem is the key ‘natural’ … it is now taken in OSM as being for both ‘natural’ (dictionary definition) and ‘unnatural’ (dictionary definition).

In an ideal world .. I would see the key ‘natural’ split into two keys - ‘landcover’ and ‘landform’ (I am not fixed on the names .. but the intention).

landcover=liquid would be suitable .. as you cannot determine what kind of liquid from a satellite view. If the type of liquid is known then a sub key can be used .. liquid=water, sewage, etc.

Comment from BushmanK on 8 April 2016 at 23:36

@Warin61, that concept you’ve mentioned is well-known and it has been discussed many times.

But it’s a kind of orthogonal to what I’m talking about now - many man-made objects, such as sewage plant reservoirs, are easily recognizable even on low resolution satellite photos, and presence of any liquid there is a kind of secondary feature of them. I mean, reservoir is always a reservoir, even if it’s currently empty (even abandoned and unusable) or if it’s supposed to be filled with some crap (literally).

Current scheme has certain false implications:

  • every kind of reservoir contains water (because water=reservoir, while landuse=reservoir and man_made=reservoir are deprecated), which is obviously false.
  • every kind of reservoir is more or less always filled, which is also false, because, for example, cooling basins are often kept empty during the cold season since air cooling is sufficient.

Again, I’m talking about completely artificial objects, while certain kinds of “man-modified” objects (such as drinking water reservoirs with natural bed, for example) could be still covered with natural= key without setting everything upside down in terms of logic and classification.

Comment from BushmanK on 8 April 2016 at 23:46

I can understand, that situation with proposal is unintentional and seems more like throwing away baby with bathwater, because author was thinking about limited spectrum of objects, but instead of deprecation of certain tags clarification of descriptions and strict demarcation of meanings should be applied.

Comment from Warin61 on 9 April 2016 at 02:16

Your talking landuse. A landuse=forest does not mean the area is always covered in trees .. for some of the time an area maybe harvested - trees cut down for processing into lumber, paper etc. I would expect the same can be implied for landuse=reservoir.

Because something is ‘depreciated’, ‘rejected’ or even ‘abandoned’ does not stop it from being used. Yesterday I used the tags “sport=cricket_nets” and “temperature=” for example. I’ll keep on doing that … untill there is something better to use, or possibly these tags are ‘recognized’. Another approach is dual tagging as existed for swimming pols where both leisure and amenity were used .. and both still exist in the database.

Comment from Tomas Straupis on 9 April 2016 at 14:16

This idea of reinventing the wheel in water tagging is totally pointless and onthologically useless. If 10 people voted “in favour” of abandoning landuse, it does not mean that we should all now change our tagging habbits. In some countries we have an agreement to mark specific objects with landuse tags and that agreement predates Zveriks proposal. This new water tagging gives absolutely no benefit I do not see why should we change our conventions and retag all objects. So we continue to tag with landuse and change newbies natural=water to landuse as well.

Comment from BushmanK on 9 April 2016 at 14:28

@Warin61, No, I’m not talking about landuse. Right now I’m talking about mapping man-made objects, which are onthologically much closer to buildings, for example.

@Tomas Straupis, I am aware that we can still use any tags we want, but approved tags, for example, have a chance to be included in editor presets, rendering rules and converter rules, while deprecated tags have much lesser chance for that. So, it isn’t only pointless, but harmful.

Comment from Tomas Straupis on 9 April 2016 at 14:50

landuse=reservoir is included in JOSM presets. And most (if not all) power mappers use JOSM.

taginfo says that 367 133 objects are tagged with landuse=reservoir, 145 633 as landuse=basin That is half a MILLION of objects :-)

And it is a known problem when anybody (even people who do more talking than mapping) can have their own “voting” and change tag status…

Comment from Tomas Straupis on 9 April 2016 at 14:56

Zverik (at least his username) is registered in 2010-05, and in 2011-03 (after 10 MONTHS of participation) he makes a proposal to change one of the most widespread tags… I remember that proposal, but at that time I skipped it because it looked pointless and ungrounded (looks terrible now).

Comment from BushmanK on 9 April 2016 at 15:15

@Tomas Straupis, OSM is growing, there are other editors and many of them coming, while developers should be able to rely on documentation, not on narrative knowledge. If something is kept only because “old mappers” still insist on using it, but documentation says different things, it could eventually fade away. This is at least inconsistent. And I’m actually more concerned about man_made=reservoir, not by landuse=reservoir, since the second one is usually assigned to hydroelectric power plant reservoirs with natural bed, while man_made=reservoir page is wiped, while this tag can represent completely different, fully artificial entities.

Comment from Tomas Straupis on 9 April 2016 at 15:35

I have nothing against new mappers and new ideas. But people should understand, that we do not start mapping each day from zero. So at least most widespread tags should not be rearanged without a very good reason and wide discussion. There are a lot of data consumers who depend on consistency of tagging. And in the case of water tagging change gives no benefit whatsoever so it should not had come through in the first place.

Comment from BushmanK on 9 April 2016 at 20:59

Honestly, this proposal reminds me so called “old lady’s order” - a concept of arranging things by size and color instead of purpose or way of using them, so it only makes everything look nice, while usage becomes awfully illogical and inconvenient.

There are also some individuals, “Wiki fundamentalists”, who always referring to documentation as to the ultimate authority, while documentation is just an explanation of idea behind tagging scheme, which could easily be far from ideal. These people making it really hard to explain other people that certain practice is bad by arguing that it’s against the Wiki articles.

Log in to leave a comment