Talk:Proposed features/reef

From OpenStreetMap Wiki
Jump to: navigation, search


There should be subtags for types of reef, and for the depth of reefs. Surface reefs and deepwater reefs should be rendered differentliy. --Skippern 17:49, 5 July 2009 (UTC)

I have added some properties in the Extended usage section. Davidbstanley 06:18, 6 July 2009 (UTC)
We could also think about a general elevation layer be it above or below sea level. If you think about the general-handling of elevation data in OSM, there are ele-tags (ele=*) for the elevation but mostly separate data-sets (SRTM) are used for hillshading and the like. I would be glad to sea sea-shading as well. -- Dieterdreist 17:32, 15 July 2009 (UTC)
One problem with using ele could be that the depth of objects in a marine situation is normally the depth below the lowest possible tide height. Whereas for a land object, it is relative to a fixed reference/datum which is not necessarily related to tide heights. There might be an argument for using both. Davidbstanley 06:57, 19 July 2009 (UTC)
ele=* and depth=* have different definition of zero (elevation above high water mark and depth under low water mark) and since the difference between these two zero values are not fixed, than they must be handled differently. --Skippern 09:45, 19 July 2009 (UTC)

Artificial Reefs

Might want to add natural=yes|no or similar, as there is a lot of sunken ships that become reefs, some were intentionally sunk to become dive areas/reefs etc. Delta foxtrot2 05:50, 15 July 2009 (UTC)

Good idea, but if not tagged, natural=yes should be implied as default value. --Skippern 10:41, 15 July 2009 (UTC)
I have added this to the proposal. Davidbstanley 07:01, 19 July 2009 (UTC)

This will need to be artificial=yes|no with no the default to prevent a problem with API0.6

Using natural for an artificial reef seems illogical at first, but if you put something underwater, it quickly becomes a part of "nature" and in general I believe reefs are considered to be in the natural(earth geography) realm. At the very least they are put there with the intention of becoming a part of the natural surrounding. artificial=yes is very vague. I would suggest tagging artificial reefs more specifically, such as: natural=reef + reef=man_made + man_made=wreck/sunken_ship/atlantis --SDT_420 07:01, 04 January 2012 (UTC)

Full Tagging

I use natural=water; sub_sea=reef; layer=-1 for the canvec2osm script. Which seems appropriate. IMO --acrosscanadatrails 05:51, 15 July 2009 (UTC) perhaps natural=reef with the following as sub-tags would be better?



reef=fringing (- Where the reef is attached to an island.)


--acrosscanadatrails 12:24, 3 March 2010 (UTC)

Alternative subsea=*

  • Thanks David(?) for taking the initiative on this and advertising it on the mailing list. Can I suggest that you/we also look at a new subsea= tag set for mapping natural (and possibly man made) underwater features that are either completely submerged or that are exposed only at low tide? I have been reasonably extensively experimentallly mapping using subsea=coral_reef for obviously coral reefs and subsea=reef for unknown and non-tropical reefs as a way or node. I have several motives for subsea=*. First, by grouping it will make searching easier. Second, we can add obscure or specialist values and if a renderer does not recognise it, it can still render out the name in blue. Third, I suspect that some day not too far away, the planet file will get so big that we wil have to look seriously about more than one database. Subsea feature is potentially a good candidate to separate out. I have not yet added many features beyond resources due to lack of out-of-copyright resources. The one exception is subsea=channel for capturing the names of navigational channels, passages, straits etc available in the Perry-Castañeda Library Map Collection ( MikeCollinson 10:03, 15 July 2009 (UTC)
I have no problem with subsea=, if it is felt that is a better option. I just want something! It would be interesting to get some feedback from others. Davidbstanley 07:00, 19 July 2009 (UTC)
  • Searching for how to tag reefs in the Maldives, where I will be on holiday in March, I found this proposal. As you're asking for feedback, here is mine:
  • +1 for subsea or sub_sea (or subwater?), for all underwater elements. I feel it more general. For example, it could also include underwater wreck (but see also historic=wreck), a non natural element. If using natural=reef with type=*, it would make more sens to use reef:type=*. --Eric S 13:02, 19 March 2010 (UTC)
  • +1 for natural=reef. Describing a reef as subsea only is incorrect. Sometimes they are, sometimes they aren't. They absolutely are natural. As far as database size, it seems like natural could be separated out as easy as subsea. Perhaps subsea could be a part of the "natural" database? Just a hunch, but I'm guessing the amount of data for things above sea level will be orders of magnitude greater than those below. --SDT_420 13:02, 04 January 2012 (UTC)

tagging as way

When adding some reefs I noticed that their tagging scheme is quite similar to natural=cliff - that is reef should be tagable as way too, not only as closed area. --katpatuka 19:18, 2 November 2009 (UTC)

Tagging as a way but not a closed area makes it dubious to render the reef. Reefs are actual areas, of varying thickness. I would think they should always be added as closed ways/areas, not just as unclosed ways. Example: Belize Barrier Reef is an unclosed way. How would this be rendered? How thick? -- Dandv (talk) 22:58, 7 June 2014 (UTC)
I agree with Katpatuka that reefs are sometimes linear features like cliffs: there is deepwater on the one side and flat water on the other side of the reef-edge. Does not make sense to map an area in such cases. RicoZ (talk) 14:13, 23 April 2015 (UTC)

Rendering support

Is there a renderer that supports the reef tag and renders it correctly? doesn't. Am I missing something?

In my opinion, the current reef tagging paradigm we are working with is inadequate to properly tag coral atolls. It may work fine for most reef tagging applications, but when we are talking about accurately creating a graphic representation of "the lay of the land" of a coral atoll, it doesn't cut it. coral atolls aren't islands where there is a landmass, a coastline and then the water. In a Coral Atoll, large areas could be considered both land and sea and must be rendered into a map view.

I am currently working on Majuro Atoll in the Marshall Islands.

If you look at the map render and then at an aerial image and you can see that a whole lot of information is missing about the "land" mass of that Coral Atoll.

On this atoll(like many others), the reef isn't some thing beyond the coastline. For all intents and purposes it is a part of the landmass. To tag it as something that disappears at the shoreline(and to not render it in a map) is incorrect in my opinion. It's more like the real coastline "land" is where the ocean meets the reef and the reef is similar to a floodplain or other "non-buildable" location.

I like natural=reef + reef=* for tagging coral atolls, I would mainly use reef=barrier, reef=outer, reef=inner.

I believe the layer=-1 tag is appropriate for many instances, but in the case of a coral atoll it's kinda blurry. I'm going to be using my OSM data to feed a flight simulator scenery generation engine(X-Plane 10) and having the reef tag with depth tags will definitely be a boon in creating more true-to-reality scenery. --SDT_420 13:02, 04 January 2012 (UTC)