I have some experience with (remote) mapping in Ethiopia based on satellite imagery, but mainly in Amhara and Oromia. I just have begun to do some mapping in parts of the Afar Region, Hari Rasu Zone (formerly Zone 5). Big parts of Afar are still very very sparsely covered in OSM.
I run into hundreds of elongated structures in rural settlements which are in all probability pastoralist or agro-pastoralist ones. These structures are in almost all cases black and white and most of them are about 2.5 x 7.5 meters big. They do not seem to be sharply angular.
For a while I thought that these might be some kind of carriages or vehicles. But this is (apart of one or two highways) definetely a region in which very few cars or trucks can be seen, if any at all.
Any hints or ideas would be very much appreciated!
I agee that - just looking on the pictures I provided - these structures look like trucks. I still think that these are definetely no trucks because of the geographical context. These are my main arguments:
Afar is one of the poorest and less developed regions of Ethiopia and the region I am mapping at the moment (Haru Rasi Zone around Dalifage) is very remote. There are just two unpaved highways which we can call highways at all, all other roads or paths can certainly only be used by 4wheel cars, if at all. I am almost sure that No trucks can drive in this region (apart the Tel highways in dry season)
The black and white structures are in almost all cases (90 %) within small settlements not close to the highways. Most of them stand in small compounds encircled by hedges or fences.
A certain portion of these structures are stable and in the same position for some years, others appear and disappear. But almost all of them are connected with small settlements or camps which have existed for a longer period of time.
The economy of this region ist still mainly a pastoralist one, based on camels, sheep snd goats. I am almost sure that 99.x % of the rural inhabitants in this region are not at all able to afford a car or even a truck.
My main hypothesis at the moment ist that these are some kind of tent-like structures.
Any ideas and hints would be very much appreciated.
They are nomadic home structures. If you do an image search for “traditional afar house” you’ll see variants. I have traveled recently in Afar and have been in Djibouti as well. These are very common in those areas. The white are usually associated with tarps or other materials to shield the occupant from wind and dust. USAID featured prominently on many of them.
They can be elongate and are generally oriented along the same axis due to local wind conditions, adjacent structures/wind breaks and sunlight.
I already for traditional Afar “mobile” homes abd have seen sone articles and pictures. But based on your posting and your experience in this region I am Happy and convinced now. Your help is very much appreciated!
Additional question: How would you tag these structures in OSM?
Good day and I’m glad I could help. Now for the next question, how to tag them?
At the risk of kicking up a storm, I looked at a few options. OSM is heavily weighted towards urban treatment of nomadic settlements. Anything outside an urban area is tourism, because why would anyone live full-time in a tent in open wilderness?!? Having lived and traveled in Kazakhstan, Qatar, Mongolia, Ethiopia etc nomadic populations with movable shelter that varies by season is not always easy to tag. I use a tiered approach where I have some knowledge of the location:
Generally I use residential: as the top level tag. From there you can further clarify as:
Irregular_settlement or informal_settlement. The definition in OSM is politely referring to homeless encampments or slums in urban areas but it also works for nomadic encampments that are seasonal. Irregular_settlement seems to be the most used so I suggest that refinement if you so choose.
Halting_site which is used to refer to designated areas for Travellers which is very Euro-centric. These have defined areas, amenities and tacit municipal endorsement.
Camp pitch or campground as an alternative although it is specific to tourism rather than a nomadic population land use classification.
I would suggest residential as an area outlining the nomadic camp including features such as water basins, wells or other amenities. Tag the amenities if they exist as separate tags. Irregular_settlement or informal_settlement can further differentiate the land use and you can always add farmyard if you can make out a holding area for livestock etc.
I agree that OSM mapping and tagging is generally heavily weighted towards urban treatment of structures including settlements and highways/paths. My experience with mapping in Ethiopia and DR Congo is that we can find a lot of formal, “regularly-shaped” settlements near the highways even in some very remote regions - but almost no mapping of settlemets in other parts of these regions in which we can only find temporary roads and paths. In some regions this means that most of the existing settlements are not mapped at all. These settlements are not in the focus of most of mappers because they follow differerent patterns then the ones we can find in Europe or the USA or in urbanized regions of Africa. The same is true with a lot of tags.
I will use “residential”, may be irregular-settlement, but not informal. I think that both irregular and informal includes a statement that these settlements are not normal or even “illegal”. I do not like this, neither in cities nor in rural areas and I definetely do not like this “undertone” when it comes to pastoralist populations.
It is great that the community is so helpful and ready to share experiences!
There is the option to propose new tags and have them added to the database. I have participated in a few, notably the naming and classification of wadi. I would support adding more clarity for seasonal/nomadic/pastoral classifications and villages. There is also a lack of detail associated with various geological features and arid environments. These arid features occur much more often in Africa & GCC regions.
I would further suggest a discussion to align on which features to propose and examples before embarking on an official proposal to the community. That way the discussion stays centered on the proposal and doesn’t get sidetracked into flagging more features outside the main goal(s).
