Talk:Proposed features/landuse=trees

From OpenStreetMap Wiki
Jump to: navigation, search

I'm still at a loss to know exactly what objective criteria you use to distinguish 'natural' from planted trees. Although in parkland, streets etc., it is possible to guess that these were planted, there are also older trees which may long pre-date the development of an area (I can think of two such groups of trees within 500 metres of where I sit). Also trees can in general manage to plant themselves quite effectively (which is why I have numerous oak & beech saplings in the garden).

Fragmenting the tagging of a fairly simple category (trees, groups of trees in rows or otherwise, and woodland (first-growth, managed, plantations, ancient, whatever) into more and more sub-categories does not help make tagging simple. Don't get so hung up about the key name "natural", it's just a useful mnemonic to help people find meaningful tags. This continual fragmentation is making OSM increasingly LESS useful for mapping trees for purposes of ecology, biodiversity, natural history and aesthetic interest. -- SK53 09:44, 18 March 2011 (UTC)

I too am puzzled at why we don't use a separate landcover tag for any wood, trees or forest, and leave landuse only to mark "purpose of area" as defined in planning documents, for example. And get rid of natural tag because it has no meaning (in most cases is equal to landcover). But for now we have to work with what we've got.
Distinguishing 'natural' from 'planted' or 'managed' is subject of landuse=forest page, but in most cases the state is fairly obvious: inside a town or along a road all trees are managed, because they are cut and disposed when they get in the way. But for trees we have only landuse=forest, and this forest word prevents mappers from using this tag more widely. This proposal doesn't introduce any new categories, it only adds a synonym to remove several ambiguities. --Zverik 10:57, 18 March 2011 (UTC)
Well I'm glad we are in agreement about a number of issues, although I'm not hugely in favour of introducing a landcover tag. I too would like to confidently tag other groups of trees (rows, small ornamental plantings, wood belts etc) which are different in character to woodland. Often the process gets tangled up in agreeing a suitable tag name: there's often substantive agreement about what we want to tag, and then the name gets in the way of agreeing how to do it.
My main point is that the managed/natural/planted distinction is mainly a huge distraction from most obvious use cases. The main ones I can think of are: navigation (it's useful to know that the path/highway goes through trees or along trees, orienteering); landcover (water run-off, solar gain, pollution ...), access (requires separate tagging, can't tell from a natural/landuse distinction), ecology (types of trees mainly), natural history (much more details of kinds of trees, plant sociology, vegetation classification, dominant species - beech wood, oak wood, larch wood ...). Use of the woodland is totally obscured by tagging everything as landuse=forest, there is a huge distinction between a commercial forestry plantation (regular felling of panels), a small poplar plantation (which will remain intact until one off felling at maturity), an ancient woodland managed by coppicing with standards, woodland managed as a nature reserve or for game. In other words if I am interested in how a tree, row of trees, group of trees, wood, forest is managed I need much more information. The other issue is that much woodland is mapped from aerial photos: how can one deduce how it is managed? I need to put these various thoughts together more coherently, will eventually blog about them. -- SK53 13:17, 18 March 2011 (UTC)