Mick,
I am a bit confused, as polar or spherical coordinate systems are
generally those making use of 1 or 2 angular coordinates. This still
has the problem of infinitesimally sized radial coordinate units at
each pole and is in fact the basis of geographic coordinate systems.
I suppose you could use alternative axes for high latitude data (e.g.
lat > 80 degrees), where the pole of the coordinate system goes
through the intersection of the prime meridian and the equator, but
that doesn't make it a universal coordinate system.
An alternative which does not have a directional bias would be a 3d
parametric cartesian coordinate system, i.e. an X-Y-Z system in which
the values are alternatively parametrized according to a geoid model.
The represents the entire earth system with equal distance precision,
but makes many earth-surface calculations somewhat non-trivial (e.g.
what compass direction is a great circle line between two cartesian
coordinate triples?). Various hybrid forms which divide the globe
into tesselated patches have also been proposed, but none have really
caught on.
Please enlighten!
-Josh Lieberman
On Aug 16, 2006, at 10:39 AM, Gregory Yetman wrote:
Hey Mick,
The short answer is legacy data and systems. We're working with
metadata and search parameters that are build on lat/long
descriptions and searches. Most of the data are in a polar
coordinate system but the metadata have been reduced to a lat/long
bounding box as required by FGDC and the NASA metadata standards.
We also want to include as much data as possible in the search
interface (distributed search), so we have to support what most
folks use in metadata -- the lat/long box.
Long term I guess that we may want to look at how ISO and Dublin
core can use alternative coordinate systems for metadata searches.
cheers,
Greg
Mick Wilson wrote:
btw, matey, I believe there are strong arguments as
why polar coordinate systems (and I mean fully 3-D
ones not constrained to an arbitrary spheroid) are the
only truly natural means for encoding geodata, one
applicable not only at the poles. You guys ever given
any thought to this, or is everyone so locked into
lat/lon that it's become the QWERTY coordinate system?
cheers
--- Gregory Yetman <[EMAIL PROTECTED]>
wrote:
Hi,
I'm working on a project that focuses on polar
metadata and we'd like to 'enable' the spatial searching for
areas that cover
the poles. To do this properly (avoid false matches), we want to do
the comparison in spherical coordinates, i.e.,
http://en.wikipedia.org/wiki/
Spherical_coordinates#Spherical_coordinates
rather than lat/long. So we are looking for software
packages (preferably open source) that will:
- transform a lat/long box into spherical
coordinates
- intersect the result with a given set of extents
and return the matches
- optionally return the area of overlap in spherical
coordinates & lat/long
I think that perhaps GeoServer/GeoTools may be able
to do this, but I couldn't determine if it would be supported
from my
admittedly quick peek at the documentation.
Any other possibilities?
Thanks,
Greg
--
--------------------------------------------------------------------
Gregory Yetman
Center for International Earth Science Information
Network (CIESIN)
Columbia University
URL: http://www.ciesin.columbia.edu/
e-mail: gyetman (at) ciesin.columbia.edu
tel: (845) 365-8982
--------------------------------------------------------------------
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
--
--------------------------------------------------------------------
Gregory Yetman
Center for International Earth Science Information Network (CIESIN)
Columbia University
URL: http://www.ciesin.columbia.edu/
e-mail: gyetman (at) ciesin.columbia.edu
tel: (845) 365-8982
--------------------------------------------------------------------
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
