[Imports] Forest Cover Data

Pavel Machek pavel at ucw.cz
Mon Oct 20 12:31:31 UTC 2014

On Mon 2014-10-20 10:53:04, SomeoneElse wrote:
> On 20/10/2014 10:12, Pavel Machek wrote:
> >Hi!
> >
> >>>I just stumbled over
> >>>http://earthenginepartners.appspot.com/science-2013-global-forest
> >>>
> >>>More info under
> >>>http://earthenginepartners.appspot.com/science-2013-global-forest/dow
> >>>nload.html
> >>Independent from the license the tree cover data is not suited for
> >>import into OSM for multiple reasons:
> >>
> >>- the data shows percentage of tree cover which is something OSM does
> >>not map.  OSM does only distinguish between forest/wood and no
> >It should be pretty safe to say that anything with >75% tree cover is
> >forest/wood.
> >
> My back garden begs to differ!

Your back garden is probably mapped already, right? :-).

> Seriously, for the reasons already mentioned it doesn't make sense to import
> this, but it does sound like it would be very useful as a way of QAing
> already imported forest / non-forest data.  It'd only be useful for _large_
> forest areas though and only as a second opinion - looking locally there are
> more than a few "false positives" and "false negatives" (showing forest loss
> when there wasn't any, and missing some forest extent / forest gain).  Also
> the data (in Derbyshire, England) seems to be offset to the south by a
> couple of tens of meters.

Maybe it would make sense to start with _large_ wood covered areas
where there's no openstreetmap data in the area. Like bigger than 2km^2.

Those are very very unlikely to be anything else than landuse=forest.
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

More information about the Imports mailing list