[Tagging] Definition of lake/pond as applied to stream/plunge pools

[Paul Allen]

On Wed, 23 Dec 2020 at 23:43, Brian M. Sperlongano <zelonewolf at gmail.com>
wrote:

> Hence my paraphrasing: streams are only modeled as ways and rivers can be
> modeled with an area.
>

Nope.  Both streams and rivers are mapped as ways.  A river (mapped as
a way) can, but need not be, enhanced with an overlaid area.  The river
is the way, not the area.  The area, if present, is likely to be many
contiguous polygons which are not part of a relation linking them
together or linking them with the river.  The polygons are visual
only.


>   So from a mapper's perspective, the difference between river/creek in
> practice isn't the successful completion of a running long jump, but a
> question of whether or not a polygon is needed to model the watercourse.
>

Still nope.  You can use waterway=river just as you use waterway=stream,
they render with different thicknesses.  There are a lot of rivers mapped
with
only waterway=river.  Both of these, where they flow through pools, have
been mapped by putting a pond there.  They were mapped that way
because, until this thread, we didn't know any better.  These were
"ponds" that had a stream flowing in and a stream flowing out
(or a river flowing in and a river flowing out).

But technically, if there is continuous flow through it, it's a pool not
a pond.  So far, so uncomplicated.  The only question is if we
bother to invent a pool tag or say that for OSM usage a pool
is tagged as a pond (as has been done many times in the
past through ignorance).

Even with a pool bulging on a river mapped as a way, it's no
different.  Map a pond, stream/river joins it at one place,
passes through, leaves at another place.

However, when it comes to wide rivers, the pool may not
bulge out.  It is an area of water that is almost still because
of the geometry of the river.  And that may or may not be a
problem with the "slap a pond on it" method.

  I assume that the vast majority of these watercourses are not mapped by
> folks testing out their track and field skills but rather are mapped from
> imagery.
>

Mostly, yes.  Over two years ago I mapped this as a pond (because back then
I knew no better and it was the only tag we had) although it's technically a
pool.  I mapped it from aerial imagery.

https://www.openstreetmap.org/?mlat=52.05532&mlon=-4.67128#map=19/52.05532/-4.67128

However, if you look around here:
https://www.openstreetmap.org/?mlat=52.08292&mlon=-4.66783#map=17/52.08292/-4.66783
You will see Netpool Slipway, Netpool Park, Netpool Wood and Netpool
Road.  They all took their name from a pool in the estuary that was
originally
used for fishing with nets and later became the hub of Cardigan's
shipbuilding industry.  If I could find any reference to the actual
extent of the pool in the river I'd map it that way (if we had a way of
doing it), but so far I've not turned up anything.

Further upstream there are many pools in the river, known to anglers
because their relatively slow flow provides good fishing spots.  You
can't map any of them from aerial imagery, only from local
knowledge.  And they're all named, because anglers refer to
them by name.

-- 
Paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20201224/56745fff/attachment-0001.htm>