[Talk-us] Wilderness Data

[stevea]

 From the USFS data mentioned earlier, I noticed two oddities that did 
not jibe with my understandings.  #1, when there is an overlap of 
boundaries between federal and state land, the two cannot be 
coincident (i.e. EITHER National Forest OR State Park must be the 
owner of the overlapped land).  #2, I thought Wilderness boundaries 
were always a "subset" of a "parent" National Forest.

I wrote and asked, and got a prompt reply from Troy Warburton, ALP 
Deputy Program Manager, USFS:

Troy-Both Forest Service and Park Service lands, excluding any 
pre-existing rights, are considered to be public lands owned by the 
United States.  The difference is administrative responsibilities. 
Areas within a National Forest Boundary may have additional 
restrictions, such as a National Park, National Monument, National 
Scenic Area etc..

Troy-A determination would have to be made on a case by case basis as 
to whether the boundaries in question are meant to be coincident but 
the Forest Service is currently in the process of reviewing and 
updating all LSRS data to insure boundaries that should be coincident 
are vertically integrated with other pertinent data sets.  As LSRS 
data is updated it will be made available through the FSGeodata 
Clearinghouse web site.

So, for #1, Federal lands can overlap other federal lands, though a 
particular distinction might have certain restrictions compared to 
another.  OK, my head wraps around that, and OSM accommodates that 
(e.g. a polygon tagged leisure=nature_reserve for Wilderness inside 
of a polygon with landuse=forest for National Forest).  Troy says 
Parks, Forests, Monuments, Scenic Areas can "blur together" like 
this, but the idea is the same for Wildernesses, as OSM tags are 
different to accommodate those.  OSM is on board:  so far, so good.

Additionally, Troy says there will be updates to these data to 
harmonize overlapping boundaries with OTHER (non-federal) 
jurisdictions, like states.  An example (the one I gave him) is how 
Limekiln State Park (California) seems to "intrude" upon what the 
USFS says is Los Padres National Forest.  I think what will happen is 
USFS will eventually get around to noticing this (these, really), 
"carve out from" its boundary for LPNF where Limekiln is, and then 
re-publish its data.  When somebody like me importing these data into 
OSM notices fresher data, I or another person will have to update, 
just like any OSM data which change over time.  In the meantime, OSM 
may display these errors in USFS's data by overlapping with other 
jurisdictions.  Yep, even the federal government admits it makes 
mistakes in data it publishes, then asks for more time to "vertically 
integrate" other data sources.  OK.

Troy-Wilderness boundaries can extend beyond a National Forest 
boundary on public lands administered by another agency.

OK, I didn't know this.  But whether one, the other, or both 
simultaneously, each are federal land, even if administered by 
different agencies (e.g. Department of Agriculture/Forest Service OR 
Department of Interior/National Park Service).  There are also STATE 
Wildernesses, for example, Anza Borrego Desert State Wilderness. 
Again, OSM does accommodate both overlapping boundaries and their 
administrative differences via tags.  These semantic distinctions 
render on mapnik as what we might call "mostly pleasing."  For 
example, both boundary=national_park + ownership=national tags are on 
USFS wilderness boundaries, but not on State Wilderness boundaries, 
which (should, but often don't) get the tag ownership=state.

However, the tag boundary=national_park is confused, as it is widely 
overused, especially on STATE parks.  Arguments are valid either way 
whether to include or exclude it on State Parks.  The reason appears 
to be that boundary=national_park is mapnik-rendered as a pleasing 
dashed green line, and name text appears at wide zoom levels (up to 
z=6).  Hence, the "overloading" of it on "non-national parks" so it 
renders anyway.

While we could quibble about whether "the fifty states are 
sovereign..." (they are) and so this means it is OK for state parks 
to have boundary=national_park tags, OR, boundary=national_park 
should be solely reserved for National Park Service "National Parks" 
(in the USA, anyway), I'd rather not.  A better discussion might be 
where as a first step we posit a new mapnik rendering rule: 
boundary=state_park renders just as boundary=national_park, except 
with dashed blue lines, instead of green.  In a new thread, please.

Oh, four National Forests -- Los Padres, Angeles, San Bernardino and 
Cleveland -- and all of "their" thirty or so Wildernesses are now in 
OSM using these fresh USFS data.  I have updated the wiki accordingly.

Thanks for reading.  Wiki pages seem too low-audience for these 
topics, talk-us seems both more-eyeballs and to the right geographic 
distribution.

SteveA
California
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20130228/06405289/attachment.html>