[Talk-us] Proposed (partial) import: New York State minor civil divisions.

[Elliott Plack]

As a fellow boundary, I say good luck to you. This is tough work 
considering how well integrated the 2008 boundaries are into the rest of 
OSM. Many boundaries overlap roads and such and so replacing them can be 
tedium. It is worth doing though, in my opinion.

On 12/15/2021 10:05 PM, Kevin Kenny wrote:
> I'll keep an eye out for places where it's right to deconflate, and for 
> opportunities to tag CDP's correctly. I'm starting with the areas of the 
> state where the original import failed altogether, and Long Island is 
> likely to come last.
> 
> A lot of the CDP's on Long Island are legal Hamlets with signed and 
> defined borders, despite having no home rule, so 'correct' tagging is 
> controversial. I don't have a good data source for Hamlets - they're 
> designated (when they are) by individual Towns.
> 
> On Wed, Dec 15, 2021 at 4:15 PM Brian M. Sperlongano 
> <zelonewolf at gmail.com 
> <mailto:zelonewolf at gmail.com>> wrote:
> 
>     As a data consumer of boundaries, this is sorely needed, and thanks
>     and sympathies for taking it on.  Will this clear up the mess of
>     improperly tagged CDPs from the original import as well as all of
>     the admin_level=8 boundaries (many of which are mis-tagged CDPs)? 
>     Long Island seems to have the worst of it, and everything is glued
>     to streets, making them exceptionally prone to breakage.
> 
>     On Wed, Dec 15, 2021 at 1:33 AM Kevin Kenny
>     <kevin.b.kenny at gmail.com
>     <mailto:kevin.b.kenny at gmail.com>> wrote:
> 
> 
>         The OSM representation of minor civil divisions in New York
>         State has several problems.
> 
>         (1) About a third of the cities, towns and villages, most of
>         them in the northern and western parts of the state, are missing
>         entirely - apparently the TIGER and USGS imports never
>         succeeded in importing them.
> 
>         (2) Many administrative boundaries in TIGER were simply
>         incorrect. The state's database appears to be more trustworthy;
>         for example, there was reasonable alignment between the state's
>         data and the county's tax rolls in the case described in
>         https://www.openstreetmap.org/user/ke9tv/diary/391486
>         <https://www.openstreetmap.org/user/ke9tv/diary/391486>.
> 
>         (3) Few, if any, of the minor civil subdivisions' populations
>         have been updated as of the 2020 census.
> 
>         (4) Some boundaries were conflated with watercourses that have
>         changed course abruptly in Hurricanes Irene (2011), Lee (2011)
>         and Sandy (2012).  Because these changes were avulsion (abrupt
>         change resulting from a specific catastrophe) rather than
>         accretion (gradual deposition of sediment), erosion, or
>         reliction (exposure of land by declining water level), they
>         often did not move the boundaries, even though the watercourses
>         have since been edited. These moved watercourses must be
>         deconflated with the boundaries, and the boundaries returned to
>         their proper places.
> 
>         New York State maintains authoritative data sets for
>         minor civil divisions at http://gis.ny.gov/civil-boundaries/
>         <http://gis.ny.gov/civil-boundaries/>.  The license for these
>         data is the state's standard open data license, available from
>         https://data.ny.gov/dataset/OPEN-NY-Terms-Of-Use/77gx-ii52
>         <https://data.ny.gov/dataset/OPEN-NY-Terms-Of-Use/77gx-ii52>.
>         This license governs several New York data products that have
>         already been imported; most notably the Department of
>         Environmental Conservation lands and campgrounds, the state
>         parks and state historic sites, and the statewide address point
>         data.
> 
>         Starting from the shapefiles available from gis.ny.gov
>         <http://gis.ny.gov>, I've succeeded in making a
>         topologically-correct OSM data set. It's available for
>         inspection at
>         https://kbk.is-a-geek.net/attachments/20211215/topo_cities_towns.osm.pbf
>         <https://kbk.is-a-geek.net/attachments/20211215/topo_cities_towns.osm.pbf>
> 
>         The tag, 'user:ke9tv:fid' present on the ways is there because
>         the conflation process that I'm using needs to know, for the
>         ways that are already done, what the unique ID of the way was in
>         the source database.  It will be removed from OSM at the
>         conclusion of the import. The other odd tags are
> 
>         gnis:feature_id - Same meaning as in TIGER: the GNIS ID of the
>         feature in question; a foreign key into many government databases.
>         nist:fips_code - The FIPS code of the feature in question;
>         another important foreign key that ties together government
>         databases.
>         us:ny:swis - The code for the feature in New York's Statewide
>         Information System, again a foreign key into many external
>         databases.
> 
> 
>         The idea with this file is that I'll go through it, town by
>         town. (There are 1600 towns; I should be able to stitch things
>         together in a few weeks.) The workflow will be to copy-and-paste
>         the relations for the municipalities into JOSM one at a time, do
>         manual conflation of the borders, do manual conflation of the
>         relations, update populations, add Wikipedia and Wikidata tags,
>         and upload. I've only done one township - Champlain, New York -
>         so far, as a test case.  It can be seen at
>         https://www.openstreetmap.org/relation/13552232
>         <https://www.openstreetmap.org/relation/13552232> .  Note that
>         the two member ways https://www.openstreetmap.org/way/1012328548
>         <https://www.openstreetmap.org/way/1012328548> and
>         https://www.openstreetmap.org/way/1012328549
>         <https://www.openstreetmap.org/way/1012328549> are already
>         split, to allow seamlessly anchoring the boundary between Chazy
>         and Mooers, the next bordering towns.  The villages of Champlain
>         and Rouses Point are also done. The latter shares boundaries
>         with the township, the county, the state and the United States.
> 
>         If there are any serious problems with this project, I'm
>         prepared to revert or amend the two changesets that have been
>         committed so far.
> 
>         Given the amount of manual work that's being done here, this
>         hardly qualifies as a 'mechanical edit' - but given that the
>         external data source is a specific file that we haven't used
>         before, I thought it wise to run this by the community.  Please
>         let me know of any objections!
>         -- 
>         73 de ke9tv/2, Kevin
>         _______________________________________________
>         Talk-us mailing list
>         Talk-us at openstreetmap.org
>         <mailto:Talk-us at openstreetmap.org>
>         https://lists.openstreetmap.org/listinfo/talk-us
>         <https://lists.openstreetmap.org/listinfo/talk-us>
> 
> 
> 
> -- 
> 73 de ke9tv/2, Kevin
> 
> _______________________________________________
> Talk-us mailing list
> Talk-us at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us