<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">I believe some waterways in Australia
will flow away from wherever the rain happens to fall ... <br>
That is a produce not just of the flatness of the terrain but also
to the quantity of rain - there can be 5 years of rainfall
delivered in a single day.<br>
<br>
Someone has put in the Australian Great Dividing Range...
fortunately it does not render as it is very rough data. And there
is no motivation to fix it .. as it does not render most are
unaware of it within OSM.<br>
<br>
On 13/09/18 20:12, Joseph Eisenberg wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAP_2vPgU7pVNU4Z=5Qfu_CO26FRsdnDVaf5O02VxePuzLyh4fA@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">Christoph,
<div>So you believe the ridges are
verifiable (and the network of waterways,
I assume), but potentially parts of the
watershed would not be verifiable because
eg. terrain is too flat? I was thinking
that in fairly flat areas it is still
possbile to see which way water flows in
drainage channels, and it's often possible
to find the highest line throught he
terrain when surveying. Also, open
topographical map sources and open source
elevation data (eg SRTM) would be the main
way to determine this. Would it be ok to
map watersheds where they are verifiable?</div>
<div><br>
</div>
<div>Would these examples be verifiable?</div>
<div><br>
</div>
<div>Wolo river is 99% surrounded by steep
ridges; good example?: <a
href="https://www.openstreetmap.org/relation/8687500#map=13/-3.8438/138.8568"
target="_blank" moz-do-not-send="true">https://www.openstreetmap.org/relation/8687500</a> see <a
href="https://www.opentopomap.org/#map=13/-3.7955/138.9242"
moz-do-not-send="true">https://www.opentopomap.org/#map=13/-3.7955/138.9242</a><br>
</div>
<div><br>
</div>
<div>
<div>The Ibele river may be questionable
in the flat valley <a
href="https://www.openstreetmap.org/relation/8687462#map=12/-4.0619/138.7478"
target="_blank" moz-do-not-send="true">https://www.openstreetmap.org/relation/8687462</a> see <a
href="https://www.opentopomap.org/#map=12/-4.0620/138.7477"
moz-do-not-send="true">https://www.opentopomap.org/#map=12/-4.0620/138.7477</a></div>
<div><br>
</div>
<div>Uwe river is mainly surrounded by
steep ridges. For the part in town
verification depends on seeing which way
water flows in open drainage ditches
along streets; <a
href="https://www.openstreetmap.org/relation/8687464#map=11/-4.2105/138.7912"
target="_blank" moz-do-not-send="true">https://www.openstreetmap.org/relation/8687464</a> see <a
href="https://www.opentopomap.org/#map=11/-4.2105/138.7910"
moz-do-not-send="true">https://www.opentopomap.org/#map=11/-4.2105/138.7910</a></div>
<div><br>
</div>
<div>Tagi river is 95 surrounded by
ridges: <a
href="https://www.openstreetmap.org/relation/8687463"
target="_blank" moz-do-not-send="true">https://www.openstreetmap.org/relation/8687463</a> ;
</div>
</div>
<div><br>
</div>
<div>Lake Habema is 98% surrounded by
ridges: <a
href="https://www.openstreetmap.org/relation/8688506"
moz-do-not-send="true">https://www.openstreetmap.org/relation/8688506</a>
; <a
href="https://www.opentopomap.org/#map=13/-4.1419/138.6722"
moz-do-not-send="true">https://www.opentopomap.org/#map=13/-4.1419/138.6722</a></div>
<div><br>
</div>
<div>
<div>I would be happy to include a warning
in the proposal and wiki that not every
watershed can be mapped in OSM. Only
those in terrain where the dividing line
is obvious and the direction of water
drainage is clear. So no watersheds
should be mapped in wetlands, flat
farmland, developed urban areas, etc.</div>
</div>
<div><br>
</div>
<div>If the issue is the lines through
flatter terrain, we could have the
watershed be a non-closed line which only
connected ridges. (Personally I feel there
are places where the watershed is obvious,
yet don't qualify as a "ridge", which
could also be included as part of the
relation). If there was no requirement to
make a closed way from the segments, this
would remove the temptation to draw
non-verifiable lines across flat land. </div>
<div><br>
</div>
<div>I value your opinion Christoph, because
I hoped this relation might encourage more
complete mapping of ridges, watersheds and
drainage basins, thus making it easier to
render good maps, eg <a
href="http://www.imagico.de/map/water_generalize_en.php"
moz-do-not-send="true">http://www.imagico.de/map/water_generalize_en.php</a></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
<br>
I note the expatiation the a river will not stop in the middle of
nowhere. <br>
In fact this does occur, a river can disappear into the sand!<br>
And some lakes have no outflow. <br>
<br>
<blockquote type="cite"
cite="mid:CAP_2vPgU7pVNU4Z=5Qfu_CO26FRsdnDVaf5O02VxePuzLyh4fA@mail.gmail.com">
<div dir="ltr"><br>
<div class="gmail_quote">
<div dir="ltr">On Thu, Sep 13, 2018 at 6:14 PM Christoph
Hormann <<a href="mailto:osm@imagico.de"
moz-do-not-send="true">osm@imagico.de</a>> wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">On
Thursday 13 September 2018, Joseph Eisenberg wrote:<br>
> Relations of type=watershed are currently used over
2000 times and<br>
> there is a descriptive Wiki page but no proposal. (<br>
> <a
href="https://wiki.openstreetmap.org/wiki/Relation:watershed"
rel="noreferrer" target="_blank" moz-do-not-send="true">https://wiki.openstreetmap.org/wiki/Relation:watershed</a>)<br>
><br>
> It would be useful to have a relation that showed
drainage divides<br>
> (aka watersheds) and drainage basins (the network of
streams and<br>
> rivers draining into a water body or waterway)<br>
<br>
Watershed divides are an abstract non-physical concept that
is generally <br>
not verifiable in practical mapping - there are cases where
they are <br>
(because they evidently coincide with physical features like
ridges) <br>
but there are huge parts of the world where they are not and
you would <br>
only try to estimate them based on already existing data.<br>
<br>
In short: This is not something you can reasonably map in
OSM.<br>
<br>
-- <br>
Christoph Hormann<br>
<a href="http://www.imagico.de/" rel="noreferrer"
target="_blank" moz-do-not-send="true">http://www.imagico.de/</a></blockquote>
</div>
</div>
</blockquote>
<p><br>
</p>
</body>
</html>