Hi GeoMechain,

The tag was changed away from landuse=residential due to complains from OSM community members that the boundaries did not actually represent residential areas. There ongoing discussions with some community members about what these areas are and whether they should be removed which you can find here: changeset/77298058

Happy to answer any further questions you may have.

Thanks,
Eli

Hi Dave,

Thanks, I have shared this changeset and the thread.

We do have a list of the areas. The easiest way to pull them out is using Overpass to query on the source and description tags across all of Thailand:

area["name:en"="Thailand"] ->.a;
way["source"~"BVBDMAY2019|WHO2019"]["description"~"."](area.a);
(._;>;);
out geom;

Note that this misses a handful of ways that were incorrectly tagged, but this should be the vast majority of them. There were also a large number of buildings that were enumerated, but I believe there were large issues with consistent tagging during the enumeration so there isn't an easy way to pull them out that I know of. I believe the organization responsible for enumeration simply looped through the ways in the query above and extracted buildings in the way boundaries which included some buildings mapped by them and some buildings mapped by other community members.

Let me know if there is anything else I can help with.

Best,
Eli

Thanks AlaskaDave and Paul_012 for your replies and the link to the forum where this is being discussed. No worries about the tone, completely understandable given that this seems to be a relatively widespread issue in OSM and the fact that this project clearly doesn't conform to OSM standards. Happy to try and rectify any damage done depending on what you all decide is best going forward.

I wanted to double check if it would be alright to share your comments and the link to the forum with the organization who made the decision to use the public instance. I believe they may be using OSM in other projects around the world and am unsure if they are consulting the community before doing so. We had previously shared with them how to set up a private instance so they could move away from the use of OSM directly, but I think it would be useful for them to be aware of the opinion of the community towards these types of enumeration.

As for the boundaries that were drawn in Thailand, they have 10 digit administrative codes because they are hand-drawn sub-villages mapped by the health centers (VBDUs) in the area. The boundaries are generally larger than the villages themselves as they were meant to encompass any potential areas of transmission (breeding sites, forests where transmission may be taking place, etc.). As far as I'm aware, there are no officially drawn boundaries for the muban/village or sub-village level in Thailand. Let me know if there is any other information I can provide about the ways that could help in deciding what the next steps are. Will keep an eye out for comments to this changeset going forward

Apologies AlaskaDave - these changesets refer to malaria foci used by the Bureau of Vector Borne Diseases in Thailand in their malaria elimination program. The description tag is a 10-digit code for identifying sub-national boundaries in Thailand's administrative hierarchy. The source tags were a unique identifier for the training in which these foci boundaries were enumerated so we could extract them more easily with the Overpass API.

The program has since moved away from using these areas given the lack of consultation with the community before deciding to use the public OSM instance. Instead we've set up a private instance to avoid polluting the public one. We have not gone back and removed these areas, but would be happy to remove the ~400 or so foci we mapped in Tak, Trat, and Ubon if you think it would be in the best interest of the community.