Mateusz Konieczny writes:
> (quoting stevea)
 "treed farmland" or "heavily wooded residential" prove slightly problematic to 
OSM tagging.

Then, Mateusz Konieczny answers:
> Map tree-covered area (landuse=forest) and map farmland (landuse=farmland) or 
> residential (landuse=residential).  Yes, the same area may be tree covered 
> and residential at the same time.


If only it were this simple, it appears not to be.  "Tree covered area" can be 
either landuse=forest (OSM's wiki defines something like a half-dozed different 
conventions on how we actually tag this) OR it can be natural=wood.  Very 
roughly stated, what _I_ do (as I see other California and USA-based users 
doing this — I'm not trying to invent a new tagging method) is to map 
distinctly "timber production" areas as landuse=forest and distinctly "appears 
to be wooded — whether pristine and ancient never-cut forest I don't 
necessarily know — as natural=wood.  That is for starters and only attempts to 
start from a point of "visible trees" (as in imagery) while only leaning in the 
direction of landuse in the aspect of landuse=forest being "it is well-known 
that this is an area which is either actively forested, or has the right to 
have its trees felled" (timber permits, owned by a logging company, CAN be cut 
but maybe are still growing to maturity, MIGHT be cut but could also be deeded 
by owner later on to become conservation or land trust protected area...).  The 
possibilities are myriad, but OSM does a "fair to good" job of characterizing 
these, and with only two tags, forest and wood.  This isn't perfect nor is the 
consensus about how we do it, so that aspect alone complicates this question, 
while at least providing SOME stability of understanding the complex semantics.

THEN there is the aspect of ALSO-has-a-residential-aspect (or perhaps PRIMARILY 
does).  Clearly, a 10 hectare / 25 acre parcel which is 98% trees and 2% house, 
garage, a small clearing and a driveway for access is something quite different 
than natural=wood (as far as its residential landuse goes).  However, it might 
not be all that different than a landuse=forest, ESPECIALLY if the residential 
land owner also has a timber permit to cut trees (possible, though not 
necessarily common, at least around here).

Regarding farmland, this has also been discussed many times, especially about 
Santa Cruz County (see that topic's wiki, the fifth paragraph of the "Work to 
be done in the County" section).  Briefly, misunderstandings happen because 
around here, we have areas which are zoned farmland, (and are actually areas of 
— among other agricultural activities — beekeeping, wild mushroom harvesting, 
herb-crafting for essential oils, other unusual but certainly agricultural 
production) but also have significant tree-cover, which may or may not be 
permitted for felling timber.  That is a whole lot of complexity to shoehorn 
into a couple-few simple tagging "rules." (Or even "guidelines").  Two 
"admonishments" in that county-level wiki are offered to prevent 
misunderstandings:  one is that "farmland isn't simply row crops" and the 
second is to read the definition of what our landuse=farmland wiki says (about 
"tillage," for example).  When both local zoning says "agricultural" and some 
activity like wildcrafting herbs to harvest essential oils both meet the 
definition of what I and others agree is "landuse=farmland," I tag these 
landuse=farmland.  These topics are complicated.  If we need more tags to 
better differentiate (I believe we do), let's coin them (with discussion and 
consensus, of course).  For example, locally, we distinguish between 
"Commercial Agricultural" (row crops), what most people would certainly agree 
is classically landuse=farmland, but we also have "Residential Agricultural," 
or what might be termed "a live-on family farm" which includes a residence / 
house and significant land, a large amount of which might be "treed," with 0% 
row crops, but allows (and actually develops) into orchards, vineyards, 
greenhouse_horticulture.  Indeed, I have tagged exactly those three latter tags 
on sub-polygons where I see them (as they are distinct tags in OSM), but in 
essence, it is 100% correct to tag the whole area landuse=farmland on the 
entire polygon (in my opinion), even though it is "also" residential.  OSM does 
not have "landuse=live-on-family-farm" as a tag, maybe we should better develop 
something like this and these.

> Yes, the same area may be tree covered and residential at the same time.

Yet, Mateusz, you don't say exactly how to tag these.  And (multi)polygons 
which describe them ARE (I know it, Doug knows it, many know it) and can be 
exceedingly complex structures to "get them right."

> Yes, "tree-covered area" meaning for landuse=forest mismatches strict meaning 
> of both landuse and forest

If only it were this simple, it appears not to be.  Again, I would go back to 
the (local? regional?) distinctions I make between natural=wood and 
landuse=forest I make above, but this is long enough already, so I'll stop here.

Steve
_______________________________________________
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Reply via email to