[Tagging] The endless debate about "landcover" as a top-level tag

Marc Gemis marc.gemis at gmail.com
Tue Jun 12 11:22:04 UTC 2018

While I'm in favour of  the landcover tag, landcover=trees is also not
really true. In most forests areas I know, trees do no occupy all the
space. There is plenty of room for grass, green plants and bushes on
the ground.
In some cases there is no other vegetation and it's just sand (or ground).

So if you use landuse=forest to indicate areas with lot's of trees,
is it then OK the overlap with other landuses ? I'm thinking of trees
in a private garden, trees in a pasture, etc.

Another problem I see are the named forests with lakes inside. Typical
mapping is multi-polygon with landuse=forest + name=xxx on the outer
ring and natural=water on the inner ring. Does this mean that the lake
is no longer part of the forest or just of the area covered with trees
? So is it also no longer part of the area named xxx ? (similar
problems with grass or bushes)

Another thing I do not understand is why we evolved from natural=wood
to landuse=forest and not from natural=heath to landuse=heath. (BTW,
natural=wood is still more popular at the moment than landuse=forest),
especially if you use the argument natural means "no human are
involved". Most heaths (if not all) are only surviving because humans
burns down parts or use sheep and cows to prevent large bushes and
trees from taking over.

How do people in GIS know how many square meter of forest there is in
a country based on OSM-data ? Is the data suited for that ?
For me a forest is a place where you can do "slow" recreation: hiking,
cycling (typically MTBs), horse riding, though I understand some
forests are purely used for the timber production
How can I find those places with OSM data ?

I thought I had an answer for all the above questions when
natural=wood, landuse=forest, landcover=trees where used "properly".
But how do you solve them in world with only "landuse=forest" ?

On Tue, Jun 12, 2018 at 12:27 PM Peter Elderson <pelderson at gmail.com> wrote:
> For landuse=forest or landuse=forestry I think landcover=trees would be implicit (default), unless another landcover is specified.
> I guess which values of landcover should be supported for rendering on OSM Carto is a matter of later discussion. For now I would be happy with grass, trees, scrub, and sand.
> 2018-06-12 11:48 GMT+02:00 Warin <61sundowner at gmail.com>:
>> On 12/06/18 19:37, Paul Allen wrote:
>> On Mon, Jun 11, 2018 at 11:41 PM, Martin Koppenhoefer <dieterdreist at gmail.com> wrote:
>>> > On 9. Jun 2018, at 15:53, Paul Allen <pla16021 at gmail.com> wrote:
>>> >
>>> > Landuse=forest could mean a group of trees which are not
>>> > consistently used by a single organization for anything (and often called "Xyz Forest"
>>> interesting, can you give a real world example where a group of trees has actually the name “... forest”? I always thought a forest would require more trees.
>> Either one of us is completely misunderstanding what the other wrote or you're quibbling about the size of a group.
>> Sherwood Forest is 450 acres of trees.  It is a nature reserve and so it is not used for forestry (aka logging).  There may
>> be occasional felling of diseased trees but it is not systematically logged on a wide scale.
>> This is why landuse=forest is problematical.  Sherwood Forest is not land used for forestry, but it is called Sherwood
>> Forest so landuse=forest may seem like the correct tag to use (because it says "forest").
>> That's why abandoning landuse=forest in favour of landcover=trees or landuse=forestry (as appropriate) is a good
>> idea.  I'll also add that I don't think landcover=trees should be used in combination with landuse=forestry because what
>> is currently on land used for forestry may not be trees but saplings or stumps.
>> I am coming around to this way of tagging.
>> Been looking at places tagged landuse=forest around me...
>> Some are forestry (yea!)
>> Some are parks ..
>> Some are nature reserves... (some of these are errors due to LPI map colours ... very similar from forestry to reserve. And yes, LPI is legally allowed in OSM)
>> Some are no more trees ... history .. though I have found one that is forestry .. just with the trees harvested and gone, they'll be back.
>> _______________________________________________
>> Tagging mailing list
>> Tagging at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
> --
> Vr gr Peter Elderson
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

More information about the Tagging mailing list