<div dir="ltr"><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial">I think this isn’t really an error. This estimation based on analyses of existing elements raba:id=1410 (currently classified as
natural=heath) – the question was – “how would OSM
mapper classify this element according to the satellite image”. More than 80%
of this elements would be classified as natural:scrub, some of them as
landuse=forest (sometimes it is very difficult to distinguish scrub from forest),
very few as landuse=meadow (1410 are former agricultural areas now overgrowth).
</span></p>

<p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial"><br></span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial">Main reason for these classification “errors”
is that the source for land cover is different (digital orthophoto 1:5000,
covering last three years – the situation might already change, detail identification
process made by experts – “normal” OSM mapper might not “see” the same).</span></p>

<p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial"><br></span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial">So, if someone would have the same knowledge
and specially the same image source, the classification error would be much lower.</span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial"><br></span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial">I think this is not the case to go for indivudual verificatilon - if it would be the case - the whole import is under question.  </span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial"><br></span></p><p class="MsoNormal"><span lang="EN-GB" style="font-family:Arial,sans-serif;background-image:initial;background-repeat:initial"><br></span></p><p class="MsoNormal"><font face="Arial, sans-serif">Martin</font></p></div><div class="gmail_extra"><br><div class="gmail_quote">On 23 November 2015 at 09:41, Martin Koppenhoefer <span dir="ltr"><<a href="mailto:dieterdreist@gmail.com" target="_blank">dieterdreist@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
<br>
sent from a phone<br>
<span class=""><br>
> Am 22.11.2015 um 23:25 schrieb colored stone <<a href="mailto:coloredstone.si@gmail.com">coloredstone.si@gmail.com</a>>:<br>
><br>
> 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.<br>
<br>
<br>
</span>if there is really an error amount close to 20% for this class you should verify all of them individually, it's definitely too high to be tolerated to just import them blindly.<br>
<br>
<br>
cheers<br>
<span class="HOEnZb"><font color="#888888">Martin </font></span></blockquote></div><br></div>