The USGS Gap Analysis Program (GAP) (http://gapanalysis.nbii.gov) has
some of the best datasets available for parks that play some part in
conservation, ball fields and the like usually won't be included in the
datasets. A GAP analysis has been completed for all states and several
regional updates have also been completed. As part of this project a
stewardship or conservation lands database would have been created.
These lands will include federal (including military bases), state,
local, and private lands that have some sort of designated conservation
status and typically utilize the best datasets available at the time.
Most states were completed in the 2000's, so they should be pretty
current. you'd be surprised how often boundaries for these lands
change. Hawaii's stewardship database was updated in 2005 and may be
the way to go for the park boundaries. I'll take a look and see if it's
decent.
As for the quality question.... it is really about scale and precision.
generally look for a scale of 1:24,000 or larger. check out the
metadata for this info. anything smaller than 1:100,000 will probably
be too coarse for osm purposes. to check precision, Tyler has the right
idea... you can see inholdings from the aerial photography and
boundaries typically follow streams and roads.
Kevin
Disclaimer: I worked on the Georgia GAP and the Southeastern GAP
projects, so I might be a little biased :)
Tyler wrote:
I did a quick web search and found that the US National Park
Service maintains a website with a large collection of GIS data,
including park boundaries and campgrounds:
http://www.nps.gov/gis/data_info/park_gisdata/hi.htm
How do people go about evaluating the quality of data that is
being considered for import? �I'm not sure there is a better
source for US National Park boundaries than the NPS itself. �It's
not like it's possible to survey the boundaries by observations
on the ground. �The only way I was able to even determine that
the previous import was problematic is that places I knew to be
just inside the park boundaries appeared just outside the
imported park boundaries.
this data isn't much better either. It may be good for some parks.
if you can verify the data is good then import.
Sometimes the USFS has better NPS boundaries than the NPS (I don't
even try to figure it out). Also, a cursory look on the wiki page for
potential data sources would have brought up both the NPS and USFS
data
sources�http://wiki.openstreetmap.org/wiki/Potential_Datasources#US_Forest_Service
I've only kind of looked at the HI state data, but there may be some
stuff in there (the large landowners and reserve maps both look pretty
coarse, though)
�
I don't mean to belabor the point, but how do people typically
evaluate the quality of data that is essentially nothing more than
imaginary lines on a map?
�
Sometimes it's easy if you know the area. "Hey, I know the park
doesn't actually include all of those houses, because they're not
inholdings" Sometimes the boundaries are "to the thalweg of Big River"
and when you see that for Big River they have what looks like a "Z"
rather than an intricate meandering river, you know they're not great
(same goes for roads, shorelines and such). The best data for any
particular boundary tends to be from the county�(in WA state)�or
property taxing authority because they greatly care who owns what, and
where so they know how much to charge in the way of taxes. Since the
state and federal government don't pay property taxes the
inter-state/federal boundaries aren't necessarily great, but the
federal/state to private landowner boundaries are fantastic. If you
really have a question, hunting down a parcel map may be the way to go.
_______________________________________________
Imports mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/imports
