[Tagging] Feature Proposal - RFC - boundary=forest(_compartment) relations

[Paul Allen]

On Wed, 10 Feb 2021 at 13:52, Martin Koppenhoefer <dieterdreist at gmail.com>
wrote:

> Am Mi., 10. Feb. 2021 um 13:47 Uhr schrieb Paul Allen <pla16021 at gmail.com
> >:
>
>>
>> The forestry proposal emphasises that its boundary may encompass
>> areas of not-trees: lakes, rocks, cleared ground where trees have
>> been felled and new trees have yet to be planted, etc.
>>
>
> I believe it is not helpful to include lakes, rocks or scree in "forestry"
> areas.
>

So what do we do about an area where we can see felled trees in aerial
imagery and which we know replanting will occur?  It is the distinction
between landcover (not trees) and landuse (forestry).


> We generalize as few as we think makes sense (and according to the time we
> dedicate), but in principle we strive for a high level of detail, because
> you can always compute more abstract generalizations, but you can never get
> detail that isn't there.
>

I agree.  But it works both ways.  If we map the boundary of forestry then
we
lose knowledge of where there are trees and where there (currently) are not.
If we map the trees we lose knowledge of where there currently no trees
but in future there will be.  Which is why I suggested in a different
response
we should map trees/not trees as at present (though possibly rationalizing
the tags used for trees) and put a boundary around the forestry area
(as we do for nature reserves, etc.)

>
> I also do not believe it is helpful to see sawmills as part of forestry.
> We do not see furniture factories as part of residential landuse, do we?
>

A sawmill isn't a landcover.  But it may be within a forestry area.  I
don't have
a strong opinion either way.

>
> On one hand
>> the cleared areas are part of forestry, and should be included in
>> the boundary.  OTOH, what is visible from a nearby hill are
>> areas of trees and areas of not-trees and in some ways it
>> might be nice to make the distinction so that people can
>> see that reality matches the map (it's always nice when we
>> can tweak reality so it matches what is on the map).
>>
>
>
> question is what you consider a "tree", does it depend on the size, or
> only the species? Is a 1 month old tree a tree?
>
> https://de.depositphotos.com/1662947/stock-photo-oak-plant-growing-in-soil.html
>

In the context of landcover, that would be scrub and might later become
trees.
In the context of a managed forest, that is still part of the forestry
boundary
(assuming it is within an area that has been felled and is not a stray
seedling
in an agricultural field or a roadside verge).

>
> If you don't expect trees that create significant shade, a cut down forest
> where new trees should grow could still be seen a forest or forestry area
> (generally I believe it is better not to clear cut all trees, if you intend
> to have a forest in the future as well).
>

Clear-cutting happens whether either of us like it or not.  The question is
whether or not that felled area will be replanted (these days it almost
certainly will be, for carbon capture).

Perhaps a more significant question is does mapping a felled area as
forestry veer too close to cadastral mapping?  Or is it a more
structured way of adding a note saying "This used to be trees, is
currently not-trees, but will be trees again"?  Land use versus
land cover.

-- 
Paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210210/7daebd3e/attachment.htm>