[Talk-us] National Forest boundaries

[Adam Franco]

Three years ago I updated the tagging and relations
<https://www.openstreetmap.org/changeset/45287531> of the Green Mountain
National Forest <https://www.openstreetmap.org/relation/2030450> in Vermont
after some discussion in the Tagging list (start
<https://lists.openstreetmap.org/pipermail/talk-us/2017-January/016986.html>,
after
<https://lists.openstreetmap.org/pipermail/talk-us/2017-January/016994.html>
some comments from Kevin).  What I ended up doing is setting the outer
"proclamation boundary" as one relation
<https://www.openstreetmap.org/relation/2030450> tagged with
boundary=national_park +
boundary_type=protected_area + protect_class=6 and the actual parcels are a
separate relation <https://www.openstreetmap.org/relation/1610352> tagged
with boundary=protected_area + protect_class=6 (and leisure=nature_reserve
for rendering -- not sure if that is still needed). Wilderness and
recreation areas within the National Forest are not members of the main
parcel relation, but instead are their own ways/relations with tagging that
indicates the higher level of protection in them such as protect_class=1b
for wilderness areas (examples: Joseph Battell Wilderness
<https://www.openstreetmap.org/way/116060596>,  Big Branch Wilderness
<https://www.openstreetmap.org/way/116060601>) and protect_class=5 for
recreation areas (example: Moosalamoo National Recreation Area
<https://www.openstreetmap.org/relation/1610350>).

I can't say that this tagging is necessarily correct, but it has proven to
be pretty useful in a few ways:

   1. The "proclamation boundary" is a big area that provides an
   appropriate name on low-zoom maps.
   2. Having the parcel relation (with cut-outs for in-holdings) is super
   useful when exploring the forest and wanting to be aware of the potential
   for no-trespassing signage.

I haven't looked at other National Forests in depth, but some in CO
(like Roosevelt
National Forest <https://www.openstreetmap.org/relation/395767> and Pike
National Forest <https://www.openstreetmap.org/relation/393066>) are just
one big relation with boundary=national_park + boundary_type=protected_area
+ protect_class=6  and no separate parcel relations. If the actual outer
"proclamation boundary" matches the main extent of the parcels that is
probably much simpler. In the case of the Green Mountain National Forest
the "proclamation boundary" almost never matches the outer edge of the
parcels, but covers a much wider area -- hence mapping both.

Hope this helps!
Adam

On Sat, Jun 20, 2020 at 11:08 PM brad <bradhaack at fastmail.com> wrote:

>
> On 6/20/20 6:19 PM, Mike Thompson wrote:
>
>
>
> On Sat, Jun 20, 2020 at 5:45 PM stevea <steveaOSM at softworkers.com> wrote:
> >
> > I think we need both as well.  I've been doing this while watching the
> evolution of how we best do this as I participate in a "do our best, always
> better" efforts to accomplish this.  Even now!
> >
> > The idea of the first kind is simply a relation with a focus on the / a
> polygon with the outer (-most) membership.  The idea of the second kind is
> one of these plus a carefully crafted inner membership, often made up of a
> complex inholding distribution containing many sometimes complex themselves
> inner polygons.
> Thanks Steve for your insightful comments.
>
> I was thinking just create separate polygons for inholdings, tagged with
> access=private and possibly ownership=private
>
> Mike
>
> I think its simpler and better to just create an inner boundary as was
> done with the Coconino NF
> _______________________________________________
> Talk-us mailing list
> Talk-us at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20200621/d3e0c313/attachment-0001.htm>