[gdal-dev] Re: gdal to define projection, EPSG:3412

Frank Warmerdam warmerdam at pobox.com
Thu Nov 12 12:38:46 EST 2009

John Callahan wrote:
> Thanks Frank/Hermann for your responses.  However, something odd seems 
> to be going on here. I feel like I'm missing something on how GDAL 
> handles coordinate system information.
> Using the same input data set, and converting to different formats, the 
> description of the output coordinate systems are different. Is this 
> expected?   Here are the gdal_translate commands and their respective 
> output files gdalinfo descriptions below.


I think the key thing is that coordinate system encoding mechanisms
tend to vary from file format to file format.  A few points of note:

  o GeoTIFF has two general ways of encoding - one is to just embed
    the EPSG PCS code for a whole coordinate system.  The other is
    to encode the projection method along with the parameters.  You
    will tend to get the first when specifying the coordinate system
    with EPSG:n and the second when using PROJ.4 strings.

  o Many geospatial formats have a particular way of representing
    projections, and may not support all projection methods, preserving
    datums, preserving names, etc.

> [1] gdal_translate -a_srs "+proj=stere +lat_0=-90 +lat_ts=-70 +lon_0=0 
> +k=1 +x_0=0 +y_0=0 +a=6378273 +b=6356889.449 +units=m +no_defs" -of 
> GTiff nt_200712_f13_v01_s.bsq nsidc_test.tif
> [2] gdal_translate -a_srs "+proj=stere +lat_0=-90 +lat_ts=-70 +lon_0=0 
> +k=1 +x_0=0 +y_0=0 +a=6378273 +b=6356889.449 +units=m +no_defs" -of HFA 
> nt_200712_f13_v01_s.bsq nsidc_test.img
> Fortunately, I can display these images fine.  They overlay nicely in 
> QGIS with some vector data I have.   However, the problem occurs when 
> projecting to EPSG:4326 (or similar) coordinate system; the output files 
> are incorrect/missing data.  (It displays data from roughly lat -39.25 
> only down to roughly -54.75 when data exists all the way down to -90.)
> One possible cause could be that gdalinfo reports (for all of these 
> datasets) latitude_of_origin as -70.  From 
> http://nsidc.org/data/atlas/epsg_3412.html, the latitude of origin 
> should be -90 and the latitude of standard parallel should be -70.  Or 
> the problem could be somewhere in gdalwarp, or specifically with the 
> Hughes 1980 ellipsoid.  I have no idea.  Thanks.

It sounds like there may be another issues with regard to the latitude
of origin in addition to the issue with GeoTIFF not working properly with
3412.  If you can isolate and file a ticket that might be helpful.

I'm sorry I haven't been following some of the recent srs threads very

Best regards,
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent

More information about the gdal-dev mailing list