[Tagging] Mapping deforestation wikipage proposal

[Paul Allen]

On Tue, 19 Mar 2019 at 10:00, Lorenzo Stucchi <lorenzostucchi95 at outlook.it>
wrote:

>
> Remember that all the ideas are related to map in the Amazonian forest an
> area where nothing is mapped and the map is empty.
>

The fact that nothing has yet been mapped there is not an excuse for
inventing new tags when
other tags exist.

The tag landcover=tree is still existing so we don’t have to propose
>

Good.  It's annoying that it doesn't yet render but it is a very good tag
where you don't know
if the trees are natural or managed.  As somebody else pointed out, much of
the "natural"
Amazonian woodland is the result of prior land management even if, today,
it is no longer
managed.  Landcover=tree is what you SHOULD use if you're mapping from
satellite
imagery and have no way of knowing if the trees are managed or not.  If you
intend to
keep track of deforestation by querying the database the fact that it
doesn't render on
standard carto isn't a problem (and if it is a problem then set up your own
tile server with
a style that does render it).

it so we will propose 3 different tag landcover = cultivation
>

We already have a tag for this: landuse=farmland (I doubt you're going to
be interested
in private gardens as counting towards cultivation, but there are tags for
those, too).  In any
case, "cultivation" isn't a COVER it's a USE.  Can you buy a bag of
"cultivation" to spread over
a patch of land?


> - barren - artificial , like for landcover = trees that is a tag for a
> wood where is not clear in which class have to be considered the wood if
> natural or maintained by humans. In such a way the other tag will be a
> higher level of tag respect to the existing ones.
>

It is unclear what you're getting at and why.  The best way to map the
absence of trees is not to
map trees there.  If you know what is actually there then map it, otherwise
don't map anything.
As somebody else pointed out from your images, one "clearing" was actually
a pond.  If
you know the lack of trees is a pond then map it as a pond, don't guess and
call it "barren" or
"artificial".  If you don't know it's a pond then don't map trees there.
It really is that simple.

We already have tags for real, VERIFIABLE objects such as quarries.  You
want a tag for
"I don't know what's there, but I can't see any trees, so I'll call it
artificial".  This is a problem
when what is really there is a pond.  We shouldn't map guesses (at least
not guesses that have
absolutely no justification).

 This doesn’t mean that we will map in a bad way or big area so with a low
> precision in the shape of the area, but we are not sure (remaining on the
> example of the wood) if the wood is maintained or not so we will use a tag
> in a more high level to don’t map wrong.
>

Hahahahaha.  Most of the objections here are because your "high-level" tags
ARE the wrong way
of doing this.

The landcover=artificial will a be a general (in the sense of the previous
> point) to map an area that is covered by buildings (residential and
> industrial area) and other man-made structures (like the quarry) this
> definition is from the nasa
> <https://ceres.larc.nasa.gov/science_information.php?page=CeresSurfID#> at
> the point 13.
>

If you know what it is (buildings, or residential area, or industrial area,
or whatever) then
map it as such.  If you don't know what it is then don't map it.  You're
calling for a tag that means
"I have no idea what this is except there are no trees" when the correct
way of handling "I don't
know" is don't map it.  In fact, you're calling for two tags that mean "I
don't know" except you
want one to mean "I don't know but it looks artificial" and "I don't know
but it doesn't look artificial"
and they are both GUESSES.  If there are no trees then don't map trees.  I
hate to have to keep
saying it, especially when so many others have already said it, but it's
that simple.

For some area where new start to watching at doesn’t exist any tag, there
> are area without trees because they are cut and there is just land without
> also grass or zone in the middle of the forest without any trees and
> nothing in some cases., I will put a photos on attach. (sorry I don’t known
> about sending photos so I put the photo on drive here the photos area1
> <https://drive.google.com/open?id=18uzLqVp1NbxZpZqhU9-Ydbc_K7i39Byl> ,
> area2 <https://drive.google.com/open?id=1BztdQZjn35sso20-mZVAWpEHs8D-ITPH>
>  and area3
> <https://drive.google.com/open?id=1kfRjnLH7lgozFKWmZIrqAu1TmX-P6kig> )
>

The best way of handling a lack of trees is not to map trees there.  That's
it.  If  you KNOW what is
present where there are no trees then map that.  Somebody else pointed out
that your guesses
about what is present in those three images was wrong because people with
more experience
than you could correctly identify what was there.

This is STILL a bad idea, on many levels.  Maybe, just maybe, it makes
sense for your particular
needs (I doubt it) but it is a bad idea for OSM.

-- 
Paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20190319/82fcec66/attachment.html>