[Tagging] Water in bays, harbors, etc.

Kevin Kenny kevin.b.kenny+osm at gmail.com
Tue Sep 6 20:56:52 UTC 2016

On Tue, Sep 6, 2016 at 3:53 PM, Christoph Hormann <chris_hormann at gmx.de>

> In other words: if you are aware of the problem you are unlikely to
> create a formal error while doing normal editing of the coastline even
> if you don't shy away from coastline edits.

I've actually learnt that, and even repaired the coastline a time or two. I
exaggerated when I said, 'stay away from the coastline,' it's more 'if you
get started editing the coastline, expect to spend an unpredictable amount
of time repairing the topology before JOSM will shut up about it..' (Which
is most likely a good thing, but it's annoying when it happens. Usually, it
means that I'm tidying someone else's mess.)

> For river mouths there is a proposal suggesting rough limits:
> http://wiki.openstreetmap.org/wiki/Proposed_Features/
> Coastline-River_transit_placement
> This more or less represents the practical mapping consensus - there are
> only a handful of larger rivers worldwide that do not comply with the
> limits suggested there.

The Hudson River is one of them, but the locals would be somewhat
astonished if the "seacoast" extended as far as
http://www.openstreetmap.org/#map=14/41.3070/-73.9655 (the typical location
of the salt front in conditions of normal flow, and a typical place above
which the river current begins to predominate), to say nothing of
http://www.openstreetmap.org/#map=16/42.7518/-73.6875 (the dam that is the
tidal limit). For what it's worth, NHD classifies that entire reach as
'estuary', which has caused rendering problems for CalTopo and TopOSM,
among others. The locals know it as the "Hudson River" all the way down to
the southern tip of Manhattan and do not think of it as an arm of the sea,
geology be damned.

> There are a number of places where people have started mapping
> significant parts of coastal waters as polygons like here:
> http://www.openstreetmap.org/relation/5405670
> http://www.openstreetmap.org/relation/3801633
> but this is not widely accepted and will cause problems for data users.

Indeed. Where I've edited, I've tried to follow local practice and make the
foreshore the multipolygon, rather than having water polygons.
http://www.openstreetmap.org/relation/4611285 is an example. (Note that
others created that particular relation; I modified it only because I
needed to reuse shoreline as the boundary of a state park. It got rather
nasty because someone ahead of me had done a forced repair to make the
coastline continuous and had broken the foreshore multipolygon in the
process. Drive-by mapping at its finest - but forgivable, because a broken
coastline is a continent-level emergency while unclosed ways in foreshore
polygons are local problems.)

> By the way you do not usually have to be that nit-picky about the exact
> definition of the water level.  Targeting the average high water line
> during the normal daily tidal cycle is usually close enough and as you
> already mentioned much better than current practice in many cases.
> Note spring tide level specifically does not mean storm water levels.
> So areas subject to storm flooding are not outside the coastline.

Right. I'm talking about neighbourhoods like
where the houses are built well above the ground on piers, and the streets
may well be awash at the new and full Moon even in good weather,
particularly if the wind is offshore. That house with the turret has a
stairway of ten or twelve steps going up to its front porch. You can see
the pilings under the house with the flagpole. (When I was growing up, it
hardly occurred to me that people in other places never needed to put on
high rubber boots to cross the street, or that raking flotsam out of the
front yard wasn't a routine chore for most!)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20160906/3bb76c13/attachment-0001.html>

More information about the Tagging mailing list