|The place tag is used on a single node at the centre of a named settlement or on an area outlining it.|
|Used on these elements|
|Documented values: 26|
|Status: de facto|
|Tools for this tag|
- Main article: Places
Used to indicate that a particular location is known by a particular name, to indicate what sort of "place" it is. A place tag should exist for every significant human settlements (city, town, suburb, etc.) and also for notable unpopulated, named places.
The simplest and most widespread way to map most place types is to position a node roughly at it's center (for a populated place having a central town hall, square or similar (far from all do), preferably there though).
If a place has a fairly well defined and verifiable outline it can make sense to map it as an Area. This allows to precisely define the extent of the place, which can be very helpful for example when using Nominatim for adress queries etc (as the results returned otherwise can often be quite confusing because of "noise" from nearby, but unrelated, place nodes ).
Note that for some applications, in particular for routing programs, not having a well placed centre node can cause problems because they'll then have to calculate a "centroid" from the area geometry, which in some cases can be very inaccurate. For smaller populated places (hamlets, villages, small towns) this is rarely a significant problem though (as most people would usually just want to be navigated to the place, not necessarily to some random mappers pick of a "center" in it).
place=island and place=islet for example typically have a well defined outline and no defined centre so they are normally mapped as areas. Very large places like place=continent, place=ocean and place=sea are by convention always mapped as nodes.
Mapping populated places as areas
Populated places (in particular place=city, place=town, place=village, place=hamlet and place=isolated_dwelling) are usually mapped as nodes since in most cases they have a well defined centre but not a verifiable outline. Some mappers map populated places as areas none the less with the geometry representing one of the following:
- The smallest administrative unit the place is part of and that fully includes the place. This often conflicts with the administratively declared places listed below and often includes areas that are not commonly considered part of the populated place.
- An aggregate/merger of urban landuses that are considered to belong to the place. This creates difficulties in particular in regions with larger agglomerations and connected settlements where it is often not clear where exactly one place starts and another one ends.
- An approximate hull polygon drawn around all parts of the populated place but not meant to represent a meaningful outline.
Because of the lack of verifiability and the different conflicting ideas for the meaning of the outline and because the well defined centre of the place which is useful and important for many applications is not part of such mapping it is not advisable for mappers to map populated places as areas but to map them as nodes. Data users should not expect the area geometries of populated places to have a particular meaning.
As of late 2017, populated places mapped as areas do not render in the openstreetmap-carto stylesheets. This is a known issue, documented at openstreetmap-carto's GitHub project. The mentioned areas does render fine in most other maps though (Osmand, Maps.me, Mapscii, the transport Map, Mapbox products, etc).
If you have a way of knowing the population of a place (from a free data source), the population=* tag typically is added to the same object the place tag appears on.
See below for the main tag values