[HOT] HOT #777 (East Cameroon) feature identification: residential or farmyard?
andrew.r.buck at gmail.com
Tue Dec 23 15:55:17 UTC 2014
-----BEGIN PGP SIGNED MESSAGE-----
Just to clarify on the buildings, they should be building=yes, not
residential since we don't know what they are used for.
Regarding the tracks, I think the animal interpretation is correct.
Areas like this often have herds of smallish animals like sheep or
goats. If the animals were left in one area they would eat too much
of the vegetation, it would all die out, and then the area would be a
desert. So the herders move them into an area for a short while, let
them nibble on what is there but not enough to kill it, and then move
them on to a new area. Hence you end up with a sort of "mesh" of main
tracks where they move from one grazing area to another and then a
fanlike pattern where the animals spread out to eat at the grazing
I agree that mapping the forested areas (these would be natural=wood I
think) is a good idea. The water and vegetation in nomadic areas like
this is the real "property". People move from one area to another so
ownership of the ground does not matter; what does matter is ownership
of the water and grazing rights and these can be fiercely contested,
often resulting in local conflicts.
Keep in mind that the description I have given above generally applies
to areas a bit drier than this (farther north in the sahel) but it
looks like a similar dynamic is playing out here as well.
On 12/23/2014 04:18 AM, m902 wrote:
> Thanks for the comments. I agree that we can't be sure that these
> are just animal shelters and so best to tag them as residential
> until someone can check locally. I'm tagging the tracks that link
> the groups of buildings with nearby settlements and anything that
> looks wide enough for a wheeled vehicle and ignoring the other
> paths. Also tagging the woodland, since the tracks don't really
> make sense without that context.. All a bit tricky because of cloud
> cover! Regards Martin
> On 23/12/2014 02:23, Blake Girardot wrote:
>> I find the whole area to be remarkable and lovely in the aerial
>> But specifically to althio's question about what makes someone
>> think they are animal tracks: Animal tracks would not have been
>> my first thought either, but once suggested I think some of the
>> paths and tracks in the area started looking a lot like animal
>> tracks to me too.
>> Here is a really nice area right around those settlements that I
>> think is covered in animal paths and tracks:
>> https://www.openstreetmap.org/#map=17/4.41335/13.89073 (Mapbox
>> There are what look like dozens of animal paths coming together
>> to go into the trees on the west side of the area where I am
>> guessing there is a watering hole, in addition to one well worn
>> highway=track (maybe) going NE'ish.
>> And 1.5km ESE of the that watering area are again, what look to
>> me like dozens of animal paths going out to a grazing area. I
>> would guess they are animal tracks because there are so many,
>> jaggy little paths, mostly faint and they seem to just sort of go
>> to nowhere, not crops or anything. But when the animal paths all
>> converge, aside from the general "jaggedness" of the path they
>> make and the nonsense layout, are almost indistinguishable from
>> man made highway=track.
>> It is a really interesting and educational area, thank you again
>> for sharing it.
>> On Sat, Dec 20, 2014 at 9:27 PM, althio forum
>> <althio.forum at gmail.com> wrote:
>>> It is interesting that you call them animal tracks. What are
>>> the hints to take them as such? My first thoughts would have
>>> been: those are regular tracks and paths used by walking
>>> humans, maybe animals and sometimes vehicles.
>>> In HOT context with remote mapping we are able to identify
>>> some agricultural fields and tag them as landuse=farmland. When
>>> the area is extensive and we can spot associated buildings we
>>> can occasionally tag the related buildings as landuse=farmyard
>>> as you suggested.
>>> At first sight the assembled buildings would be settlements in
>>> my mind so I would tag landuse=residential. But I am very
>>> attentive to anything you could add to identify this area as
>>> animal tracks and farmyard.
>>> Is it easily possible to identify pasture/meadow from imagery ?
>>> For very detailed rural mapping we could then tag this as
>>> landuse=farmland or even landuse=meadow.
>>> _______________________________________________ HOT mailing
>>> list HOT at openstreetmap.org
> _______________________________________________ HOT mailing list
> HOT at openstreetmap.org https://lists.openstreetmap.org/listinfo/hot
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
-----END PGP SIGNATURE-----
More information about the HOT