I’ve mapped some intermittent waterways (i.e. oueds, wadis, dry rivers …) in the central Sahara (Ahaggar region, mainly) and left most intersections with highways untagged. In the larger part of these cases, this is intentional on my part – because I mostly can’t be more specific based on Bing imagery:

fords

Some intersections of intermittent waterways and highways have been tagged as fords (e.g. here). Some of these seem plausible to me, generally where the waterway may be classified as a stream and little water flow may be expected (e.g. here).

However, where the waterway is wider, in many cases I don’t think one can make a safe judgement from aerial imagery alone whether a given intersection will be fordable during wet season or not. Images of the dry rivers in their peak flooded state might be a start, but they seem rare in the Bing imagery.

From knowledge on the ground, I can’t recommend to assume by default that a waterway–highway intersection is a ford. People drown in places like that.

bridges

At many of these intersections, there is no particular built infrastructure. Many or most earth roads simply cross the dry river bed without change in surface or structure. After a rain and flooding, these crossings may be damaged and be repaired or perhaps even relocated.

On major highways, there may be built infrastructure. In the region I’m interested in, this is probably rather the exception than the rule. In a few places, there appear to be bridges which are discernible by their shadow in Bing imagery (e.g. here).

In other places, the road is built to be flood-resistant (e.g. on concrete foundations) – without necessarily being usable during flooding, I’d judge.

Other highway sections look (in Bing imagery, to me) like they might be slightly elevated on what might be called an embankment (e.g. here) – though from the imagery I’d judge that they are not fortified in any other way and will probably also be flooded and may then be unusable (and possibly damaged). I think there might also be culverts in some places, but I find that difficult to discern in Bing imagery.

Personally, I don’t always feel confidently able to tell from the Bing imagery what kind of built infrastructure is present at a waterway–highway intersection where there’s a built road.

conclusions

With surveys during (several consecutive) wet seasons one might be able to accurately classify crossings as safely fordable during normal rains, or possibly as flood prone during heavy, exceptional rain.

My default assumption will remain that unfortified earth roads crossing even moderately wide dry rivers in the central Sahara will be unusable during flooding.