#18297 closed enhancement (fixed)
Invalidate natural=water with no other tags
Reported by: | pangoSE | Owned by: | Don-vip |
---|---|---|---|
Priority: | normal | Milestone: | 19.11 |
Component: | Core validator | Version: | |
Keywords: | natural water | Cc: |
Description (last modified by )
Hi
I think lakes only tagged with natural=water is wrong.
Could we add this to the validator?
In Västernorrland in Sweden alone I found ~1700 lakes missing the water=lake tag.
(FYI I did this:
natural=water and key!=waterway and key!=water in "västernorrland" <- overpass turbo wiz
"natural=water and -water=* and -waterway=* and (type:way or type:relation)" <- search in josm - I excluded nodes to avoid some incorrectly tagged bays near Sundsvall)
Attachments (0)
Change History (15)
comment:1 by , 5 years ago
Description: | modified (diff) |
---|
comment:2 by , 5 years ago
Description: | modified (diff) |
---|
comment:3 by , 5 years ago
Description: | modified (diff) |
---|
comment:4 by , 5 years ago
Description: | modified (diff) |
---|
comment:5 by , 5 years ago
Keywords: | natural water added |
---|---|
Milestone: | → 19.11 |
Owner: | changed from | to
Status: | new → assigned |
comment:6 by , 5 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
comment:8 by , 5 years ago
Replying to Wulfmorn:
But not all water is a lake...
There are other values of the key water available. It only warns if the key water is not set at all.
comment:9 by , 5 years ago
Sorry, I was a bit quick because i had to run out to lunch :) I mean, this issue is about Sweden (the example), and in Sweden a small body of water is not a lake, its a "tjärn", which I believe is not a pond. You COULD map everything as lake, but I've always felt that there is something missing - same as with river/stream/ditch.
Anyway, I think this validation should also warn about too big ponds and too small lakes.
comment:10 by , 5 years ago
@wulfmorn I disagree. The definition of a lake in osm is not determined by its size but by whether it is man made or not and whether it is open to the sea or not, etc. If you want all tjärn in sweden I suggest setting up a postgis environment and search by size or name.
The limit for when a tjärn becomes a lake is completely arbitrary and prone to subjective differences.
comment:11 by , 5 years ago
This issue has nothing to do with lakes. As said we only check that natural=water has a water key. For tagging discussions, please don't use the JOSM bugtracker.
comment:12 by , 5 years ago
See here for further discussion related to water tagging: https://www.openstreetmap.org/user/pangoSE/diary/391133
follow-up: 14 comment:13 by , 5 years ago
Uhm... Doesn't it make mapping more complicated? Like, why not now warn about natural=wood without leaf_type, or highway=* without surface?
comment:14 by , 5 years ago
Replying to Zverikk:
Uhm... Doesn't it make mapping more complicated? Like, why not now warn about natural=wood without leaf_type, or highway=* without surface?
I think not. Both your examples are hard to discern from orthophotos and thus makes it harder. In a swedish forest the leaf types are often mixed and I personally don't thing it provides so much value to add this as it changes every 50 years because of forestry and replantation.
It is often not hard to distinguish a reservoir as defined by the wiki from a lake. Natural=water can be literally anything from a little fountain to a reservoir for a dam and thus do not contribute much information in itself which leads to lower data quality.
On the other hand we have highway=road which we could also invalidate I think as it suffers the same problems (could be anything from a trunk to a footway).
In 15512/josm: