GALEON folk -
A word of caution regarding use of coordinate reference system designators:
EPSG, following conventional usage for "geographic" coordinate reference
systems, specifies that the order of components in the coordinate string is
Latitude-Longitude. This contrasts with (most) projected coordinate systems which use
easting-northing (though some also do use westings and southings!).
In the fragments shown below I see that the coordinate reference system is EPSG
4326 applying to coordinates like (-70.599602,41.550406) If this is intended to
designate a place just off Enderby Land near the Antarctic, then everything is
fine. However, I suspect the intention is in the North Atlantic, in which case
it is wrong.
This is a well known issue, that has led to period outbreaks of heated
discussions on the OGC lists. By this stage it is not productive to just
object, proclaim stupidity on the part of the mainstream, and go your own way -
that is not a path to interoperability.
In order to resolve the issue, OGC has issued the designator
"urn:ogc:def:crs:OGC:1.3:CRS84" to refer to the same CRS, but with longitude
before latitude:
see OGC document 05-010
https://portal.opengeospatial.org/files/?artifact_id=8814
and the revisions proposed in 06-023
http://portal.opengeospatial.org/files/?artifact_id=13937&version=1
Suggest that you correct either the coordinate reference system designator, or
the coordinate strings.
However, if my assumption regarding the inended location is incorrect, my
apologies and ignore this message.
Simon Cox
----- Original Message -----
From: "Rich Signell" <rsignell@xxxxxxxx>
To: <galeon@xxxxxxxxxxxxxxxx>
Sent: Saturday, June 03, 2006 4:21 AM
Subject: Using "double" for lon,lat for GALEON WCS NetCDF3 files
GALEON-folk,
I think it might be a good idea to use type "double" for "lon" and "lat"
when requesting "format=NetCDF3" from the Galeon WCS.
I used the remote data serving capability of the GALEON WCS to grab a
piece of a CF-compliant NetCDF file on my own OpenDAP server.
I asked for both format=NetCDF3 and format=GeoTIFF.
I then put the resulting NetCDF file on my OpenDAP site and used GALEON
WCS to grab another piece of data as GeoTIFF.
The pixel resolution on the resulting GeoTIFFs should be the same, but
it is different, because the original NetCDF had type "double" and this
is 3 arc second data from the NGDC Coastal Relief Model.
[rsignell@ricsigdtlx bathy]$ gdalinfo file1.tif
Driver: GTiff/GeoTIFF
Size is 541, 421
Coordinate System is:
GEOGCS["WGS 84",
DATUM["WGS_1984",
SPHEROID["WGS 84",6378137,298.2572235629972,
AUTHORITY["EPSG","7030"]],
AUTHORITY["EPSG","6326"]],
PRIMEM["Greenwich",0],
UNIT["degree",0.0174532925199433],
AUTHORITY["EPSG","4326"]]
Origin = (-70.599602,41.550406)
Pixel Size = (0.00083333,-0.00083333)
FILE2:
[rsignell@ricsigdtlx bathy]$ gdalinfo file2.tif
Driver: GTiff/GeoTIFF
Size is 541, 421
Coordinate System is:
GEOGCS["WGS 84",
DATUM["WGS_1984",
SPHEROID["WGS 84",6378137,298.2572235629972,
AUTHORITY["EPSG","7030"]],
AUTHORITY["EPSG","6326"]],
PRIMEM["Greenwich",0],
UNIT["degree",0.0174532925199433],
AUTHORITY["EPSG","4326"]]
Origin = (-70.599602,41.550407)
Pixel Size = (0.00083160,-0.00083542)
It's only a fraction of a meter, but it's still a bit unsettling to see
the pixel resolution go from uniform to non-uniform.
Here's the script I used:
#!/bin/bash
# Example script to grab gridded Coastal Relief Model data via the
GALEON WCS
# Get Vineyard Sound bathy from *my* CF-Compliant, DODS-served
# Coastal Relief Model data via GALEON. COOL!!!!
# format=GeoTIFFfloat
curl -o file1.tif
'http://motherlode.ucar.edu:8080/thredds/wcs/?dataset=http://stellwagen.er.usgs.gov/cgi-bin/nph-dods/rps/data/crm_geo.nc&request=GetCoverage&version=1.0.0&service=WCS&format=GeoTIFFfloat&coverage=z&bbox=-70.6,41.2,-70.15,41.55'
# format=NetCDF3
curl -o file1.nc
'http://motherlode.ucar.edu:8080/thredds/wcs/?dataset=http://stellwagen.er.usgs.gov/cgi-bin/nph-dods/rps/data/crm_geo.nc&request=GetCoverage&version=1.0.0&service=WCS&format=NetCDF3&coverage=z&bbox=-70.6,41.2,-70.15,41.55'