[Tagging] Double, double, toil and trouble (how to map Witch's Cauldron?)

[Mateusz Konieczny]

I would tag it as a waterway in tunnel (though I have no idea about a suitable value) oras water area with covered=yes and natural=bare_rock area mapped, both with a proper tags.
It also sounds like it is a tourism=attraction .

Sounds a bit similar to https://en.wikipedia.org/wiki/Blue_Hole_(Red_Sea) <https://en.wikipedia.org/wiki/Blue_Hole_(Red_Sea)> - thoughthis one is entirely underwater and of a different origin.

7. Oct 2018 19:11 by pla16021 at gmail.com <mailto:pla16021 at gmail.com>:


> I've encountered a feature called, in English, "Witch's Cauldron" (also "Witches Cauldron" and "Witch's Pit") and> called, in Welsh, "Pwll y Wrach."  It was mapped by somebody else around 4 years ago and the mapping has> one definite error and a couple of things that may be wrong.  The problem is I'm not sure what the correct> mapping would be.
> It's here: > https://www.openstreetmap.org/node/3200691239#map=18/52.07127/-4.77079 <https://www.openstreetmap.org/node/3200691239#map=18/52.07127/-4.77079https://www.openstreetmap.org/node/3200691239#map=18/52.07127/-4.77079>
> It's a complicated geological feature.  I've yet to come across any description of this particular feature that> gives a name for that type of feature, and perhaps there isn't one.  The only way I can describe it is by the> processes that led to it.  It once consisted of a stratum of soft rock at sea level, overlain by harder rock.> Tidal erosion formed a cave by removing the soft rock, leaving a layer of hard rock forming the roof.  Eventually,> part of the roof collapsed.  The result is a hole in the ground with sea water at the bottom, with a tunnel from 
> the hole to the sea.  Depending upon the state of the tide it's possible to traverse the tunnel from the sea to> the hole (but only in something like a kayak, nothing larger).
> It's been mapped as an area of natural=water (no other tags).  In the centre of the water is a node tagged> natural=arch, which is not an arch at all.  The arch is about 25m NW of that node.  OS OpenData StreetView> (available as background imagery in iD and possibly in other editors) shows a thick, grey dashed line> connecting the water in the hole to the coast's high water mark and nearby are the words "Natural Arch."
> It currently shows a tributary of nearby river connecting the hole in the ground.  Such an interpretation is> not backed up by the OS or ESRI backgrounds (Bing is too unclear to cast any light on the issue).  Nor> is this backed up by any description of the feature I've found.
>
> So natural=arch is in the wrong place.  Arguably it should be a closed way covering the water passage 
> underneath, possibly with layer=1.  The tributary that isn't visible in OS should be removed.  And some sort> of water should be mapped (whether it renders or not, just for routeing) under the arch connecting sea to>  hole in the ground.  But what sort of water?  And what additional tag to use for the water in the hole?  It's> not really a pond, it's part of the sea.
>
> I can't think of a good way to do it.  The least bad train of thought I had was how it would be mapped if the> arch collapsed.  In that case the HWM would extend inland to encompass the hole in the ground, which might> perhaps be tagged as a cove, because that's what it would be.  So why not do that with a natural=arch over it?
> Any better ideas?  If nobody can come up with anything convincing, I'll leave it alone and pretend I never saw> it. :)
> -- 
> Paul
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20181007/3fcc2424/attachment.html>