[Talk-us] Potential data source: New York City watershed recreation lands

Kevin Kenny kevin.b.kenny at gmail.com
Wed May 25 02:22:09 UTC 2016


On 05/24/2016 08:28 AM, Elliott Plack wrote:
> Kevin,
>
> Nice work on the proposal. Scraping geo-data from PDF is a feat.
> Kudos! I've read through your proposal and having worked on a
> statewide protected area import, I support this one.
> protect_object=water is a novel one. Are there other uses of this tag,
> e.g. protect_object=trees for state forest?
>
> Cadastral data is always tough to manage in OSM, especially in
> backwoodsy areas where parcels are only one or two steps removed from
> the original subdivision by the King of England back in the 1600s or
> earlier. Much of Maryland's state forest boundaries are what I'd call
> a "best guess" in GIS. On the ground, they are usually marked.
> especially near populated areas, so they are observable. With this,
> you'd be open to locals finding the boundary to be incorrect and then
> updating, correct?

Of course! What I'm NOT open to is having an armchair mapper go tracing 
some third-party map and clobbering the state's data. The state's GIS 
data are actually getting to be pretty good. There have been boundary 
monitors locating witness trees with GPS, and there has also been some 
recent interagency coordination so that the NYS and NYC parcels are 
starting actually to appear to be in alignment.

And this is New York. The land divisions were from Holland, not England. 
Except where they were from England. Or France. Or two or more of the 
above, in conflict with one another. When doing cadastral analysis in 
this part of the world, I'm forever referring back to 
http://www.hardenbergh.org/patent/CatskillLandMap1970.pdf. (For example, 
one of my brother's property lines in Tusten Township is the boundary 
between Great Lots 9 and 10 of Division No. 1 of the Minisink Patent... 
and thus specified on the deed. That line was fixed before the French 
and Indian War.) Any property line that doesn't follow a lot boundary on 
that map is automatically suspect.

There is good signage along the highways and trails. A lot of the rest 
of the parcel boundaries, as I said before, are marked simply with paint 
or hatchet blazes, or with corner monuments (cairns and/or surveyors' 
pins). There are also often old fence and field lines that can be spotted.

I realize that 'protect_object' was misspelt: it should be 
'protection_object'. I plan to fill in many of the remaining fields from 
http://wiki.openstreetmap.org/wiki/Tag:boundary%3Dprotected_area .



-- 
73 de ke9tv/2, Kevin




More information about the Talk-us mailing list