[Imports] Slovenia landcover import RABA-KGZ review

[colored stone]

Sorry for not answering for so long time.


Only recently we have a meeting of a small local group of OSM mappers that
are preparing a land cover data import. We have analysed the “problematic”
tags (natural=heath, natural=moor and natural=bare_rock).


We have been looking to already imported data and we have generally agreed
that existing tagging might be problematic. We have concluded that
natural=scrub corresponds to raba:id=1410  much closely. Not really exact –
but the estimation is than more than 80% of existing elements of
raba:id=1410 could be classified as natural=scrub. We have concluded that
these elements should be imported with tag natural=scrub.


Situation with raba:id=5000 and raba:id=6000 is much more confusing. We
have been agreed that these two elements should not be imported.


Prior to importing new areas existing data imported will be improved:

-       - Elements with raba:id=5000 and 6000 will be deleted

- -    -  Elements with raba:id=1410 will be changed to natural=scrub

We are intending to use JOSM with query operation (like:
*user:the_one_who_import
and “raba:id” =1410*) for each area (tile) already imported. If there is
any better technique to perform such operation, please let us know.


Best regards,


Martin



On 12 July 2015 at 18:36, Christoph Hormann <chris_hormann at gmx.de> wrote:

> On Sunday 12 July 2015, colored stone wrote:
> > Please find clarifications/explanations on the Slovene agricultural
> > land use (shortly RABA-KGZ) to OSM tag translations. For the
> > clarifications we have mainly used the “methodological paper on
> > agricultural land use” or “the interpretation rules” available at
> > http://rkg.gov.si/GERK/documents/RABA_IntKljuc_20131009.pdf.
> > [...]
>
> Thanks for the additional information.  I won't argue the points based
> on the methodological paper in detail since i can not read it in the
> original language and i am not sufficiently familiar with the data to
> judge how well the actual data complies with the specifications.
>
> A few general points though:
>
> - use of landcover tags, in particular things like natural=scrub and
> natural=heath in OSM is frequently quite inprecise.  This is not a good
> reason to be less strict with tagging in an import.
> - the OSM landcover tagging unlike the classification scheme of your
> source data is not a closed system.  Not every area on earth matches
> one of the OSM tags.
> - individual positive examples do not mean much - no one questions that
> each of the source data classes also contains areas that match the
> planned OSM tags.  The question is how much of the data does *not*
> match the planned tagging.
>
> Therefore here a few examples from the already imported data of what
> does not match your planned tagging:
>
> natural=heath:
>
> http://www.openstreetmap.org/way/353487056
> http://www.openstreetmap.org/way/353231651
> http://www.openstreetmap.org/relation/5271375
>
> As far as i can see most of these are grassland (or other herbaceous
> vegetation like ferns as well as Blackberries, other Rosaceae and
> similar plants) with a varying amount of larger scrubs and mostly
> smaller trees.  In many cases this is land that is in the process of
> being reclaimed by trees - either previously cut forest areas or former
> farmland no more used.  natural=heath specifically means dwarf scrub
> vegetation (that is species that naturally do not grow tall, not young
> trees).  In Slovenia's climate this is rare outside high mountain
> areas.
>
> Interestingly most areas that would qualify as natural=heath are
> probably included in class 5000.
>
> natural=moor (to be natural=fell):
>
> http://www.openstreetmap.org/way/353333307
> http://www.openstreetmap.org/relation/5274011
> as well as all other areas with this tag except for the Alps - none of
> this is close to the alpine tree line so natural=fell does not match.
> Use in the Alps is questionable as well but since natural=fell is a
> fairly vague tag you can't really say it is wrong.
>
> natural=bare_rock:
>
> http://www.openstreetmap.org/way/351883914
> that is a great example for natural=scree - and here it is specifically
> excluding the rock area above
>
> http://www.openstreetmap.org/relation/5252002
> mixture of bare_rock, scree and sparsely vegetated areas
>
> http://www.openstreetmap.org/way/353153325
> that would be natural=shingle
>
>
> --
> Christoph Hormann
> http://www.imagico.de/
>
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/imports
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20151122/ef9cd264/attachment.html>