[Talk-us] Potential data source - NY DEC points of interest

Kevin Kenny kevin.b.kenny+osm at gmail.com
Fri Dec 2 20:54:14 UTC 2016


On Fri, Dec 2, 2016 at 2:40 PM, Adam Franco <adamfranco at gmail.com> wrote:

> Hello all,
>
> After doing a bit of hiking and canoeing in the Adirondacks this summer I
> noticed that OSM has very poor coverage of campsites & lean-tos, something
> I'd find particularly useful. The New York DEC POI shapefile
> <http://gis.ny.gov/gisdata/inventories/details.cfm?DSID=1258> has the
> locations and details of these available and regularly updated.
>
> Before going down the road of putting together an import proposal I wanted
> to find out if anyone has information about usage grants for the DEC data
> sets. From following this list I've heard that Kevin Kenny has imported
> other DEC lands shapefiles for the park and area boundaries:
> https://lists.openstreetmap.org/pipermail/talk-us/2016-May/016290.html
>
> Do you Kevin [or anyone else] know if permission to use DEC lands also
> applies to other DEC data sets or if additional permission would be needed?
>
> I've added an entry about this data souce to the OSM Wiki's Potential
> Datasources page
> <https://wiki.openstreetmap.org/wiki/Potential_Datasources#New_York> and
> would like to update it with more concrete information either from
> usage-grants already received or by making a new request to use the data.
>
> Any info or feedback on this datasource would be appreciated.
>

Thanks for taking an interest in New York! We surely have a massive mapping
problem/opportunity here: a park that's larger than some European
countries, with a population of only 130,000, most of whom have no Internet
access. Hard to get it mapped, to say the least!

Please do NOT import the DEC PoI file; not from a legal perspective, but
from a data quality perspective.

I do use it as a data source in my personal map because of spotty OSM
coverage (as you observe), but as OSM has been improving, I'm about to
discard it. A typical example of what I'm talking about is visible at
http://kbk.is-a-geek.net/catskills/test3.html?la=42.0966&lo=-74.5095&z=14 -
showing Rider Hollow in the central Catskills. I show point-of-interest
icons from both OSM and the DEC PoI file, and trails from both OSM and the
DEC Roads and Trails file.

The 'parking' icon at the end of the county road from Mapledale is in OSM
and is correctly spotted.

The 'campsite' icon that underlies it is data from the PoI file. I've been
in there a few times and never spotted a designated campsite anywhere near
the trailhead. There's a private house no more than fifty feet outside the
state's property line, and DEC surely would not have placed a campsite so
close to it (ordinarily they're at least a quarter mile from habitation).
There's probably a marked campsite in there somewhere. I've never tried too
hard to find it. There are a couple of splendid spots for 'at large'
camping in among the hemlocks up in Mine Hollow, and I've just stayed up
there.

The 'information' icon refers to the trail register, comes from the PoI
file and appears to be correctly spotted.

The lean-to and associated privy are within sight of the bridge, are from
OSM, and look to be right. The outline from the building import that Tom
Hynes did http://www.openstreetmap.org/way/304800815  is kind of
questionable - either it's some distance from the lean-to, or it represents
some other building that's long gone.

The parking icon to the south of the lean-to and privy at the 2400 foot
level is from the PoI file and is obviously wrong. I suspect it is supposed
to represent the parking where Rider Hollow Road enters the Big Indian
Wilderness and is digitized from a map at an inappropriate scale (1:250000
or even smaller). There's a similar problem with the Lost Clove parking
areas on the NE flank of Balsam Mountain.

The lean-to at the confluence of the three streams is from the PoI file.
The file names it, so I know it's supposed to be the one by the bridge and
is placed on the wrong stream junction.

The trails shown in magenta are clearly misaligned. OSM has them right
according to my GPS. DEC Roads and Trails has them wrong. The Pine
Hill-West Branch trail definitely follows the ridge through the col between
Balsam and Haynes Mountains. The Mine Hollow trail makes a wide switchback
away from the stream to where it joins the Pine Hill-West Branch, which
does fall off the ridge to the west in the area between Balsam and Belle
Ayre Mountains.

Not shown on my map, but present in OSM, is a spring
http://www.openstreetmap.org/node/2845338153 . The PoI file has the spring
also, incorrectly sited on the east side of the ridge.

I see similar issues in any area where I have (literal!) 'boots on the
ground' knowledge of the terrain. In my opinion, the file simply isn't
accurate enough to bother with an import. I've been trying, instead, to use
it as a 'to do' list - identify PoIs that I want to be sure are in OSM if I
happen to pass them with a GPS. Lately, in the Catskills, others have been
beating me to it. Given that I'm just getting back into hiking after over a
year off with a foot problem, and that I'm starting slow so that I don't
just wind up back in rehab, I've been focusing on easier hiking in the
Helderbergs, trying to fill in the gaps in New York Long Path and do some
more of the little nature preserves along the escarpment.

It could be that there are regions that the DEC has paid better attention
to - everyone knows that DEC tends to neglect the Catsills. I noticed that
the agreement with GPS was much better when I thru-hiked the
Northville-Placid Trail and collected GPS tracks. The major disagreements
were the relocations at the southern end and the places where our friend
Castor canadensis has decided that the trail is interfering with one of his
projects. The worst data problems there were missing campsites and
shelters. I'd need to see more comparison with data in the field to be
convinced. Did you collect tracks and waypoints on your outings that we can
compare with the DEC data?

As far as the legalities go, I was relying on the original permission that
Russ Nelson got to import the DEC Lands data. I haven't reviewed them in
detail myself, since all I was doing was to repeat an earlier import to
tidy up some problems.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20161202/17a4f8cf/attachment.html>


More information about the Talk-us mailing list