I’ve been working on a proposal to document surface=laterite as a dedicated OSM surface value for roads, tracks, and paths surfaced with lateritic soil: the red/orange, cohesive tropical soil that is firm and dusty in dry season and plastic/sticky or impassable when wet.
Currently, mappers fall back to surface=dirt, surface=ground, or sometimes surface=clay for lack of a dedicated value. A bot migration stripped hundreds of segments including major highways in Cameroon and Gabon of any surface tag by converting surface=laterite to material=laterite without adding a fallback.
The proposal is in draft and I am specifically looking for feedback from mappers who have surveyed roads in sub-Saharan Africa, West Africa, or other tropical regions before moving to RfC. Your field experience carries more weight here than any desk analysis.
A few things I’d welcome input on:
Does the field description match what you see on the ground?
Are there regional names or local soil terms worth documenting?
Are the scope boundaries (what’s in, what’s out) practical from a mapping perspective?
This looks familiar, in Ghana it is called laterite too and as you are already aware this is not easy to tell from aerial imagery. Based on my knowledge from Ghana and neighboring West African countries, there are roads in their natural forms that exhibit same properties but one can not say it’s literite hence use of surface=unpaved or surface=ground
I personally have never used surface=clay in Ghana because where there is laterite it is mostly put to form hence surface=compacted
Are you trying to account for wet/dry with surface=laterite since the material changes in both states?
I also see a lot sub-tags and uncertainties packed into surface=laterite that might make it difficult to interpret. Even though a road could be in good condition and compacted, because the material is laterite, contributors might be forced to using surface=laterite , I can see you have tried to make most of these cases clear, which makes it overloaded, and of course it’s OSM tags they can be complex.
Good to hear the term is locally recognised. And yes, surface=unpaved or surface=ground is exactly the right fallback when you can’t confirm the material from imagery alone, the proposal keeps those as the default for remote mapping.
Your practice is correct. There are two distinct cases: natural laterite soil road = surface=laterite; laterite gravel quarried, imported, and engineered into a road surface = surface=compacted. No conflict between the two, the material origin and processing is what separates them.
Yes, and that’s a core reason for the tag. Laterite has a known seasonal pattern, near-frictionless on first rain, then deeply plastic, that surface=dirt doesn’t capture. The tag is a material value, but the material carries that predictable wet-season risk as part of its identity.
Fair point. The proposal is long because it needs to address every adjacent tag to survive the RFC process, not because the tag itself is hard to apply. A streamlined version with detail moved to annexes is in progress.
Please do @Enock4seth , and anyone else mapping in Africa is welcome to weigh in too.
Worth clarifying the distinction: all the laterite road photos I found on Wikimedia Commons show natural fine-grained laterite soil, lightly graded or compacted by traffic only. That surface still turns to plastic red “soap” in the rainy season regardless of how firm it looks in dry season, and would qualify as surface=laterite.
The engineered version used on some urban roads in Thailand contains around 35% gravel-sized laterite concretions, is machine-rolled, and sometimes has fine gravel added on top. That surface is more rain-resistant and falls under surface=compacted. It is also less common locally since most of the terrain already sits on natural laterite soil.
I’d like to recommend that you prepare some material for StreetComplete support:
a list of countries where this surface should be offered as an option
the localized name of the surface in each relevant country-language pair
It seems like you have most of this information in the proposal page somewhere, just collect it in one spot for easy use by the developers.
I suppose one unresolved question is if it’s appropriate to offer surveyors both “Laterite” and “Compacted laterite” (surface=compactedmaterial=laterite) side by side. (In which case, a compacted laterite representative photo is very important.)
The dry photos show a fine-grained red-orange surface with scattered gravel-sized particles, consistent with the natural concretions that occur in laterite profiles. The wet photos show water ponding flat across the full width rather than draining, which is the fine clay fraction sealing the surface. That a surface with visible gravel-sized material still behaves this way in rain is actually useful evidence: the coarse particles are embedded in a cohesive matrix, not forming a free-draining skeleton the way a true compacted gravel surface would.
But your point about graded and rolled roads is a real problem I don’t have a clean answer to yet. If a road has been graded by machinery but the underlying material is still dominant fine-grained laterite soil, does it still turn plastic in the wet season? From the photos it looks like it might. If so, surface=compacted overstates the rain-resistance, but surface=laterite may feel wrong to mappers who watched a grader improve it.
That boundary is something the proposal needs to address before the RFC. Your field experience is exactly what would help settle it.
