Overall goals

The Hadjer Lamis area is very poor, and there is an unusual burden of disease and malnutrition amongst the population which contributes to high mortality in children under five years old. In order to better understand, assess, and respond to this, we need to know more about the population.

We are mapping villages and taking their names on the ground, but identifying all of the inhabited areas and counting the structures is much more efficient from aerial photos. Knowing where all of the villages are scattered through the savanna helps us to map them, and counting the buildings within each village gives us a quick and fairly accurate method to estimate population (important to understand the spread of disease and identify areas of highest need). Perhaps surprisingly, structure counts are often more accurate than asking how many people live in the villages.

The tasks at hand are:

• Find all of the villages in the area, and draw an area around them, tagging each as Land use, Residential.

• Map the structures in each village. This is done by tracing all buildings as polygons and tagging them as Building Features=Building (in iD editor) or building=yes (in JOSM) , or by simply counting them and adding a tag Structures with the appropriate number.

Village structure

Settlements in this part of the world are generally organised into extended family compounds, each containing a number of small shelters and often a few crops. You can see the outlines of the compounds as dark lines dividing the village in to smaller sections. The villages are generally more or less circular.

This village (Absoufa) is quite typical of the villages in the area. Here’s what a little part of it looks like on the ground.

Villages much like this are scattered all over the landscape in Hadjer Lamis. It is difficult to find them all on the ground, as even the local people do not necessarily know all of the locations of all of the villages! If they are already traced, it is much easier for our field teams to find the villages to tag them with the appropriate local names and other information.

Try to trace around all of the compounds, not just the houses; the compounds often extend well past the inner circle of houses but are definitely part of the village.

Houses

There are two basic types of houses here: round huts, often called Tukuls, and rectangular buildings. All are made of some combination of mud and straw (in some cases the straw has been pre-processed by donkeys or cows, making it stickier).

The tukuls are easy to spot from the aerial imagery. They are round, with a distinctive central point where the roof peaks. They look a little bit like, well, um, a rounded mound with a darker-coloured point at the top centre… you know, like a mushroom cap.

Please simply tag all structures as “building”. There is a distinction between “house” and other types of building, but it’s incredibly difficult to tell the difference from an aerial photo, so best to simply leave them as “building” for now.

The rectangular houses may have a thatched roof or a corrugated metal roof. A metal roof is usually quite obvious from the aerial photo as it reflects a lot of light. In general, the metal-roofed buildings are more likely to be important buildings such as a mosque, chief’s home/office, or health clinic. If you see an obvious metal roof, you can add a tag for roof:material - metal.

Some of the buildings may be for storage of grains or animals (i.e. chicken coops). You can’t tell from the aerial photo, so don’t worry about it! Just trace all of the standing buildings you can see. The primary goal of the building tracing is to estimate population, and we can do this simply by multiplying the number of structures or roof area by a constant determined by surveys we do on the ground.

It is fairly common to see a faint ring or rectangle where a building once existed (mud and thatch homes do not last forever). It is best to count only currently standing structure, as this is most likely to correspond to the actual population. However, in the event that you can’t tell whether the building is still there or not, please trace it! When in doubt, add the feature. It’s easier to remove it than to find a missed one.

Compounds

Unlike areas further south in Africa, where villages usually consist of a collection of homes without much internal separation, villages in Hadjer Lamis are basically organised by family compounds. Each compound is surrounded by a fence of mud, wood, or brush, and contains several buildings (sometimes including grain storage buildings as well as houses). Many compounds also include some cropland, in what is effectively a small farm or large garden.

These fences really define the internal structure of the villages.

Sometimes buildings are integrated into the compound wall; often you’ll see that a rectangular building shares an outside wall with the compound perimeter.

There is no existing default tag for “compound” in OpenStreetMap. Therefore, when they are mapped as an Area and simply tagged “compound”, they do not appear in the OSM map (though they are present in the data). For this reason, until we are able to resolve this, we are not tracing the compounds in the current task. However, please learn to recognise the compounds as they are part of the village; if you only trace around the houses the village perimeter will appear smaller than it actually is.

Overall goals

The Hadjer Lamis area is very poor, and there is an unusual burden of disease and malnutrition amongst the population which contributes to high mortality in children under five years old. In order to better understand, assess, and respond to this, we need to know more about the population.

We are mapping villages and taking their names on the ground, but identifying all of the inhabited areas and counting the structures is much more efficient from aerial photos. Knowing where all of the villages are scattered through the savanna helps us to map them, and counting the buildings within each village gives us a quick and fairly accurate method to estimate population (important to understand the spread of disease and identify areas of highest need). Perhaps surprisingly, structure counts are often more accurate than asking how many people live in the villages.

The tasks at hand are:

• Find all of the villages in the area, and draw an area around them, tagging each as Land use, Residential.

• Count the structures in each village. This can either be by tracing all of them as polygons and tagging them as Building, or by simply counting them and adding a tag Structures with the appropriate number.

• Trace the tracks and/or roads between the villages.

Village structure

Settlements in this part of the world are generally organised into extended family compounds, each containing a number of small shelters and often a few crops. You can see the outlines of the compounds as dark lines dividing the village in to smaller sections. The villages are generally more or less circular.

This village (Absoufa) is quite typical of the villages in the area. Here’s what a little part of it looks like on the ground.

Villages much like this are scattered all over the landscape in Hadjer Lamis. It is difficult to find them all on the ground, as even the local people do not necessarily know all of the locations of all of the villages! If they are already traced, it is much easier for our field teams to find the villages to tag them with the appropriate local names and other information.

Try to trace around all of the compounds, not just the houses; the compounds often extend well past the inner circle of houses but are definitely part of the village.

Counting versus tracing houses

It is much faster to simply count every house than to trace them all. However, it is rather difficult to count accurately, especially when it is not immediately obvious which buildings exist and which are shadows, rocks, wells, trees, or other features. Furthermore, it is hard to verify that the count is correct (verification takes at least as much time as counting in the first place), and it is hard to maintain consistency between mappers (some people may count a ring-like structure which may be an old, collapsed house, while others may exclude these from the count). If all buildings/structures are traced, validators can more easily see what is consistently counted or not.

If you plan to simply count the structures, please add the following tag to the village polygon: Structures - 42 (or however many structures you counted). Do so in ID editor simply by clicking on the village polygon, hitting the + button in the ‘all tags’ section on the left, typing Structures into the left-hand box and the appropriate number into the right-hand box.

If, on the other hand, you are willing and able to trace all of the structures, don’t worry about counting them, we can do so automatically from the tracing you’ve done!

Houses

There are two basic types of houses here: round huts, often called Tukuls, and rectangular buildings. All are made of some combination of mud and straw (in some cases the straw has been pre-processed by donkeys or cows, making it stickier).

The tukuls are easy to spot from the aerial imagery. They are round, with a distinctive central point where the roof peaks. They look a little bit like, well, um, a rounded mound with a darker-coloured point at the top centre… you know, like a mushroom cap.

Please simply tag all structures as “building”. There is a distinction between “house” and other types of building, but it’s incredibly difficult to tell the difference from an aerial photo, so best to simply leave them as “building” for now.

The rectangular houses may have a thatched roof or a corrugated metal roof. A metal roof is usually quite obvious from the aerial photo as it reflects a lot of light. In general, the metal-roofed buildings are more likely to be important buildings such as a mosque, chief’s home/office, or health clinic. If you see an obvious metal roof, you can add a tag for roof:material - metal.

Some of the buildings may be for storage of grains or animals (i.e. chicken coops). You can’t tell from the aerial photo, so don’t worry about it! Just trace all of the standing buildings you can see. The primary goal of the building tracing is to estimate population, and we can do this simply by multiplying the number of structures or roof area by a constant determined by surveys we do on the ground.

It is fairly common to see a faint ring or rectangle where a building once existed (mud and thatch homes do not last forever). It is best to count only currently standing structure, as this is most likely to correspond to the actual population. However, in the event that you can’t tell whether the building is still there or not, please trace it! When in doubt, add the feature. It’s easier to remove it than to find a missed one.

Compounds

Unlike areas further south in Africa, where villages usually consist of a collection of homes without much internal separation, villages in Hadjer Lamis are basically organised by family compounds. Each compound is surrounded by a fence of mud, wood, or brush, and contains several buildings (sometimes including grain storage buildings as well as houses). Many compounds also include some cropland, in what is effectively a small farm or large garden.

These fences really define the internal structure of the villages.

Sometimes buildings are integrated into the compound wall; often you’ll see that a rectangular building shares an outside wall with the compound perimeter.

There is no existing default tag for “compound” in OpenStreetMap. Therefore, when they are mapped as an Area and simply tagged “compound”, they do not appear in the OSM map (though they are present in the data). For this reason, until we are able to resolve this, we are not tracing the compounds in the current task. However, please learn to recognise the compounds as they are part of the village; if you only trace around the houses the village perimeter will appear smaller than it actually is.

Roads/tracks

There are very few roads in this area that resemble Western-style roads. While there is one main paved highway through the region, almost all roads are dirt tracks. For that matter, most of them are temporary; during the dry season vehicles can pass more or less anywhere and don’t need to follow a specific road. Whichever route is most used during the dry season then tends not to grow many plants during the rainy season, so it becomes the default road for that rainy season! As a result, you’ll often see a bunch of intertwining paths between major villages (especially if the imagery is taken during the dry season), where each car has taken its own slight variation on the route.

So how do you define a road as a line on a map when there’s no single permanent road?

This requires a certain amount of judgement. The first thing, obviously, is to see where there are actual tracks. Once you’ve found a track, look to see if there are other tracks nearby, especially larger ones (or groups of tracks) that may be more likely to be the “official” road. Finally, zoom out, take a look around, and see if the road actually goes somewhere! A very good clue that a track, or set of tracks, is actually an often-travelled road is if it connects two settlements, or connects a village to a seasonal watercourse (where there may be a well where people fetch water). If, on the other hand, a track goes out from a village and ends in a field somewhere, it’s quite likely that it’s nothing more than the fresh track of a single donkey cart whose owner travelled to their field on the day the satellite photo was taken.

A rough guide to road classification levels:

• if it’s paved, it’s a primary road. In fact, it’s pretty much national highway #1.

• If it’s a clearly defined single road with permanent-looking sides, and goes between large population centres (i.e. from a town to a large village), it’s a secondary road.

• If it’s a multi-tracked road that obviously sees a lot of passage, and goes between villages, it’s a tertiary road.

• If it’s a single track, but clearly goes somewhere interesting (like another village, or an obvious water point, or to a larger road), it’s a track.

• If it’s a single track and doesn’t obviously go anywhere interesting, it’s (at best) a Path, but more likely it’s just where somebody happened to go with their donkey cart recently; don’t bother tracing it.

• A path within a village, particularly if it obviously links compounds or large features like water points, or connects to a tertiary road, it might be worth tracing and labelling Path. However, use your judgement in doing this; while it might make the map of a given village a bit more clear it doesn’t serve much purpose in terms of identifying population or facilitating humanitarian work.

asdf