[Imports] Slovenia landcover import RABA-KGZ review
chris_hormann at gmx.de
Fri Jun 19 16:56:45 UTC 2015
On Friday 19 June 2015, Stefan Baebler wrote:
> A co-importer just fixed the import wiki page (
>KGZ ) and we're double checking every landuse type tags to make sure
> the wiki is correct.
> After wiki page is fixed, we'll let you know and ask for another
> review. Of course we will fix the already imported data, if needed.
Thanks. Based on what is in the wiki i can already say that several of
the feature classes do not work well as you translated them to OSM
tags. Here the Google translations of the descriptions of those i
raba:id=1410 => natural=heath:
> The land, which are growing due to the omission of farming or modest
> agricultural use. On it appear the young woody or thorny vegetation,
> and trees and shrubs, usually of different ages whose cover is
raba:id=1500 => natural=scrub:
> The area overgrown with trees and shrubs, which cover more than 75%
> and not classified in the woods. I'm also includes waterside
> overgrowth, if riverine belts overgrown with trees and bushes, and
> hedges of forest trees and shrubs.
natural=heath/scrub are defined by height, not by percentage of
coverage. These classes essentially represent woodland or scrubland of
varying density. A grassland with scattered trees (which falls under
1410 i suppose) is not natural=heath.
raba:id=5000 => natural=moor
> Non-forest land covered with low vegetation (less than 2 m), which is
> barren or inaccessible. The coverage with vegetation is no more than
natural=moor - despite being more or less deprecated - does not match
that description. natural=fell could be used in some cases (i.e. in
the alps above the tree line) but does not fit for areas free of trees
for other reasons than due to altitude. Usually these areas will be
either natural=heath or natural=grassland.
raba:id=6000 => natural=bare_rock
> Undeveloped land with little or no vegetation, making such areas can
> not be included in any other class. This includes all land covered
> with bare rocks, sandy beaches and dunes, pebbly surface on or in
> water courses, scree and other open spaces
This applies to all non-vegetated areas, not only bare bedrock. This
includes natual=sand, possibly natural=beach, natural=scree,
natural=shingle and possibly also non-natural bare ground areas.
Tagging all of this natural=bare_rock is not a good idea.
Similar problems occured with past landcover data imports - the usual
classifications in these data sets differ from those used in OSM making
a 1:1 translation impossible.
More information about the Imports