search

jackba / arctos

Automatically exported from code.google.com/p/arctos
0 stars 0 forks source link

Higher Geography clean-up and service #268

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
Higher Geography needs consistency and can be automated, since it's an
arbitrary value that can be generated independently in data entry. 

First quick fix: Add higher geography values for data entry to pick from
based on GADM's list of resources in these combos:

1) continent/ocean
2) continent/ocean: country
3) continent/ocean: country: state_prov
4) continent/ocean: country: state_prov: county (for US only?)

That should cover 90% of higher geography needs for now.

To  implement, create a list of the values needed (MK is working on),
create a list of current HG from respective collections for
clean-up/match-up. A little db work and we can start using a HG list that
we all agree on. This would not require collections without resources to do
data-cleanup to do anything except continue with their own "authority" but
future data-entry can start to use a conventional HG (and spelling, for
instance).

Future better fix: would be an HG lookup service where a coordinate pair
would spatially resolve HG, lat/long, UTM coordinates, PLSS, USGS Topoquad,
or anything else we want to add (just keep adding polygons). (See limited
services like Geographical Locator, MSU). Possibly would be supplied in
BerkeleyMapper (actually would make a lot of sense there) and pass back the
values to Arctos (or temp tables, if necessary). Think the reverse BG.

Original issue reported on code.google.com by koomap...@gmail.com on 2 May 2009 at 1:20

GoogleCodeExporter commented 9 years ago

Original comment by dust...@gmail.com on 13 May 2009 at 5:51

GoogleCodeExporter commented 9 years ago 
AC - where are we going with this? In the new (2012) model (pending licensing) 
we can look up and store (as a string) geog (in Locality) and continue to use 
geog_auth_rec as we have been. Geog_auth_rec still needs massive cleanup to 
support data cleaning services, but perhaps we could strive for exactly one way 
to represent the textual description of a point.

Original comment by dust...@gmail.com on 3 Jul 2012 at 3:03