[Tagging] Feature Proposal - RFC - boundary=forestry(_compartment) relations (Was "Feature Proposal - RFC - boundary=forest(_compartment) relations")

[Kevin Kenny]

On Sat, Feb 13, 2021 at 11:09 AM Paul Allen <pla16021 at gmail.com> wrote:

> On Sat, 13 Feb 2021 at 15:50, David Marchal via Tagging <
> tagging at openstreetmap.org> wrote:
>
>> In such case, the forestry area would me mapped as a dedicated feature,
>> separately of the surrounding protected area. The proposal includes an
>> example tagging for such cases in US National Forests (
>> https://wiki.openstreetmap.org/wiki/Proposed_features/boundary%3Dforestry(_compartment)_relations#US_National_Forest);
>> is your example significantly different and in need of a dedicated tagging
>> example?
>>
>> Pembrokeshire Coast National Park is not a national forest.  As I said, it
> includes hamlets, villages and towns.  It also includes coastline (the
> hint is in the name) and beaches.  There are even industrial areas
> that are totally unconnected with forestry, trees or any kind of wood.
> And it has some areas of managed forest within it.
>
> This isn't a case of trees-within-trees, as your examples are.  It's
> a case of a managed forest inside a protected area that is not
> even close to being all trees.  There are other types of protected
> area than woodland, and those protected area may have managed
> forests within them.
>

Yes, indeed.  The Adirondack and Catskill Parks in New York are similar.
(Note that they don't have 'State' or 'National' in their names: they're
_sui generis_ legal entities separate from New York's system of State
Parks, and the Federal government has nothing to do with their creation.)

To complicate matters further, about half their land area is designated as
Wilderness, Wild Forest, Canoe Area, or Primitive Area. (All of these are
roughly protect_class=1b; the fine details of the regulatory distinctions
are immaterial here.) Perhaps two-thirds of the remaining land is actually
used for productive forestry, by companies like International Paper - and
most of the rest, outside the villages and the few industrial sites, is
tree-covered. (Moreover, the conservation easements that bind some of the
timber companies allow for public access to company land to a limited
extent.) What we're talking here is a nearly contiguous swathe of trees
covering a land area comparable to that of Belgium, with some tracts
devoted to forestry (sensu lato - managed to preserve the trees), some
devoted to forestry (sensu stricto - managed to extract products from the
woodland), and some not legally bound to either (but tree-covered, and on
which forestry is likely practiced privately, to a purpose unknown to OSM.)

My principal interest is in mapping the regulatory structure. That governs
where I can go, what permission, if any, I need to go there, who provides
emergency services, and what I may do when I get there. The lands that I'm
interested in mapping are signposted, at least along the highways. A lot of
the boundaries in the back country haven't been surveyed and marked in the
field - and that includes everything up to and including the township and
county lines. Nobody cares, within reason, if I trespass across an unmarked
and unsurveyed border. People do indeed care if I camp too close to water,
or in the subalpine or alpine ecozones (defined locally as being above 3500
or 4000 feet elevation respectively), or on what is unambiguously priivate
land. People do care if I cut living wood in a wilderness area, or if I
hike off trail in a timber company's holding.

The trees don't respect these boundaries, There are ponds, marshes,
avalanche scars, and patches of alpine tundra in land of all
classifications. The fact that a given patch of land is or is not
tree-covered is nearly independent of whether that land is in a managed
forest.

While the designation of purpose - land use - does follow cadastre, land
ownership is also less than informative.  There is producing forest owned
by the state, and there are what are effectively wilderness areas owned and
operated by NGO's or other private concerns such as hunting clubs.

I see a lot of the disagreement here as an attempt to fit these
cross-cutting concerns into a hierarchy. Some argue that the 'top-level'
classification ought to be the landcover ("there are trees here", vs "there
is something else here").  Others say it's the regulatory regime. ("people
may hike here", "the trees here are protected from harvest", .....) Still
others say that it's the ownership ("this land belongs to the state", "this
land belongs to a timber company", "this land is NGO owned and maintained",
...)  All of the hierarchs contend as a result that whatever attributes
that cut across the concern that they are trying to address are therefore
secondary in some sense.

Another recurring contention is that tagging these objects falls under 'we
don't map cadastre'.  I'm not interested particularly in mapping cadastre -
except to the extent that public access and use constraints follow it.
Which, in the US, they do, of course.  A private landowner has
near-unlimited say in who may and may not enter upon his demesne, for what
purposes and under what circumstances.  Nevertheless, there are many such
landowners (NGO's, universities, municipalities owning extraterritorial
land, ...) that consider public education and recreation to be among their
missions and therefore allow for public access. This is the extent to which
I think we should map cadastre. I'd have no intention of importing my
county's tax maps willy-nilly, even if the county did not (rightfully or
wrongfully) claim copyright in them.

The situation of mulltiple conflicting hierarchies has led to what seem to
me to be ridiculous results in which I could not map a state forest as a
unified whole despite its surrounding signage, unified name, and unified
management plan - because it divides into areas of woodland combined with
areas of other things such as ponds and meadows. (The argument: "if it's
not tree-shaded, it can't be part of a forest!")  This situation was
mitigated somewhat by the introduction of protected areas - which, I remind
you, took nearly a decade from proposal to rendering and remain
controversial to this day.  Still, there remain non-government managed
forest lands that are still dedicated to forestry but, because of harvests,
beavers or avalanches, might not be tree-covered at any given time.  These
are not protected areas in any current sense of the term, but they are in
some sense 'forest' areas.

I'm not saying that the proposal at hand is perfect. I'm merely making the
plea to consider that there is not perfect overlap among any two of the
concerns of land use, land ownership, and land cover.  I welcome any
resolution to the situation that doesn't run afoul of misinformation, which
is what I see in the arguments that trumped "no aliases" (I've advanced the
argument for why what's proposed is not an alias for anything else), or "we
don't map cadastre" (we do map the boundaries of public recreational
features!), or "this is already covered by existing tagging" (in six
competing views, all of which either fail actually to addresses the
concern. or else take on a misleading appearance in most renderings.
Existing tagging appears to offer no correct means for tagging these
features.)

I particularly urge people to consider their vote carefully if they've done
most of their mapping in a part of the world that does not have large
contiguous stretches of forest. In places where the native forests went,
centuries ago, in to the hungry maws of fireplaces, stoves, tanneries,
forges and furnaces, forests are designated, protected, well-circumscribed
things. I ask those for whom that's the only familiar view of a 'forest' at
least to acknowledge the judgment of those who inhabit wilder and less
organized places. We're trying to map what we have, as impossible as what
we have seems to those in more organized communities and nations.
-- 
73 de ke9tv/2, Kevin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210213/0ff63010/attachment-0001.htm>