[Imports] Slovenia landcover import RABA-KGZ review

colored stone coloredstone.si at gmail.com
Sun Jul 12 09:16:42 UTC 2015

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

This methodological paper explains elements of RABA-KGZ in more details
(this is a practical manual how to map agricultural land use). In some
cases definitions slightly differs from definitions from Regulations
(sub-law) which are included in RABA-KGZ wiki page.

We have included some partial translations of definitions and illustrations

*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*

>* 20-75%.*

*OSM Tag:natural=heath*


A dwarf-shrub habitat, characterised by open, low growing woody vegetation,
often dominated by plants of the Ericaceae (see

According to the methodological paper (especially examples) this category -
raba:id=1410 (*agricultural land overgrowth*) – is very close to the
definition of tag:natural=heath.

Examples of areas belonging to this category (page 57):
·       dense area of trees and shrubs along the forest / forest edge,
which according to the Forest Service are not included under forest,
·       neglected permanent crops, which is visible process of overgrowth,
·       fern area where there are visible signs of overgrowth (young shrubs
and other woody vegetation)
·       areas where they removed trees, shrubs and other overgrowth (eg.
the forest edges, forest edge, embankment, etc.), if the result of modest
or inadequate agricultural use again overgrown,
·       areas overgrown with young pine trees that have mushroomed on the
land for the abandonment of agricultural land use
·       Kanela (special habitats plain in Primorska region).

Examples on page 61 (*slika=picture* 79 and 80): Agricultural land

Example on page 62 (picture 81): Cleared area at the edge of the forest,
which is overgrown again

Example on page 62 (picture 82): Agricultural land abandoned as a result of
the abandonment of agricultural land use (grazing,

mowing, plowing) or insufficient load of animals in the area

Example on page 63 (picture 83): Areas covered with blackberry are
classified as 1410

Example on page 63 (picture 84): Areas overgrown with young pine trees as a
result of the abandonment of agricultural land use are classified as 1410


Some practical examples from existing OSM map shows that such areas are
mostly categorized as *natural=scrub*. Rarely as *natural=heath*, sometimes
even *landuse=forest*. Sometimes it is difficult to distinguish between
*natural=scrub* and *natural=heath*. In some cases the interpretation from
original sources for RABA-KGZ and OSM will always be different - original
source for RABA-KGZ is digital orthophoto 1:5000 not older than 1 year at
the time RABA-KGZ is captured.


Our proposal is to leave the exiting translation *raba:id=1410 =>
natural=heath* as it is. Optionally this category might be translated as
*natural=scrub* as well. But then we are losing the differences between
these two classes of RABA-KGZ. The definition also fits quite well and
later changes by local mappers can always be applied if the situation 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.

The percentage of coverage is not a main distinguishing parameter between
tags like natural=heath, natural=scrub and landuse=forest. According to
definitions and examples from methodological paper the translation from
*raba:id=1500* (named: “trees and shrubs”) to *natural=scrub* shouldn’t be
problematic. The definition is very similar to OSM definition. Confirmed
also by analyse of existing natural=scrub in OSM map comparing to RABA-KGZ

Example on page 69 (picture 92): The area covered with trees (this is not

Example on page 69 (picture 93): The area overgrown with trees and shrubs
(this is raba:id=1500)

Example on page 70 (picture 94): The area overgrown with trees and shrubs

Example on page 70 (picture 95): Terraces on the lawn, overgrown with trees
and shrubs

Example on page 71 (picture 96): Example of tree in a row – not classified
as raba:id=1500

*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*

>* 75%.*

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.

Agree - natural=fell is better as natural=moor.


*Habitat above tree line in alpine and subpolar regions, principally
covered with uncultivated grass, low growing shrubs and mosses and
sometimes grazed.*

Example on page 104 (picture 144): Inaccessible land

Example on page 105 (picture 145): Barren land

Example on page 106 (picture 146): Barren land with no soil cover

Example on page 106 (picture 147): Barren land

Example on page 107 (picture 148): Dry open land with special vegetation

*Solution:* existing elements created by RABA-KGZ imports raba:id=5000 will
be replaced to natural=fell. *The methodology to be prepared.*

*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.


*An area of bare rock is sparsely vegetated or not vegetated at all, so
that the solid bedrock becomes visible.*

*Areas to tag could be*

·       *high altitude mountain areas where the rock lays bare, like
summits, aretes and rock faces, but not scree, shingle or moraines.*

·       *rocky coasts*

·       *rocky areas in uplands, desert and polar regions*

·       *outcrops like open pit mines or road cuts*

·       *monolithic mountains like Ayers Rock.*

Short analyses of import elements shows that imported elements raba:id=6000
covers this definition almost completely.

There is an additional explanation in the methodological paper:

*Despite the general description (identical to the description in the
Regulations) the actual land use 'Dry open land with negligible or no
vegetation cover" does not include:*

*• Periodically dry river courses,*

*• Sandy beaches and dunes,*

*• Gravelly areas along streams or in.*

Example on page 109 (picture 150): Open land without, or with insignificant
vegetation cover

*Prepared by coloredstone*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20150712/fb479082/attachment-0001.html>

More information about the Imports mailing list