[Gdal-dev] Re: GDAL and Spot Vegetation
neteler at itc.it
Wed Aug 18 15:08:10 EDT 2004
On Wed, Aug 18, 2004 at 01:07:10PM -0400, Frank Warmerdam wrote:
> Markus Neteler wrote:
> >On Wed, Aug 18, 2004 at 12:28:35PM -0400, Norman Vine wrote:
> >>Vincent Schut writes:
> >>>About Plate Caree, I'm not sure either. I used a simple lat/long wgs84
> >>>configuration which looked OK. It is however possible that better
> >>>projection parameters exist...
> >>"Plate Caree" is just another name for "Equidistant Cylindrical"
> >>i.e. linear scaling of latitudes and longitudes
> >>In other words, all meridians and parallels are straight lines.
> >How should I write that as SRS OGC WKT?
> >While I have addressed the wgs84 things, the projection info is somewhat
> >missing. So far I didn't find any example.
> Norman is right about Plate Caree. It is also known as Equirectangular in
> some circles. The page I maintain about it is at:
No doubts (I was aware of both the name - looking into the Evenden docs
as well as the url).
But I wasn't aware of 'testepsg' - thanks for the hint. Exactly what
I was looking for.
> The "testepsg" program from gdal/ogr says:
> warmerda at gdal2200% testepsg '+proj=eqc +lat_ts=0 +lon_0=0 +x_0=0 +y_0=0'
> GEOGCS["WGS 84",
[for other users a note:
cd ogr/ ; make testepsg
> A recent update (just last week) in the VRT format driver allows use of any
> "user input" for describing coordinates systems. That means you should
> recent CVS builds) be able to just use the PROJ.4 string to describe an SRS
> if you want. If not, just embed the WKT with all the yucky XML escaping.
> Actually, come to think of it, there likely aren't any XML escapes actually
> required for WKT in VRT XML though by default VRTs produced by GDAL will
> all possible XML escape sequences .. not just the required minimum. Now I
> am getting off topic. :-)
..."use the PROJ.4 string to describe an SRS"... excellent!
So no more excuses to write the proper VRT file :-)
Thanks for all your helpful comments.
More information about the Gdal-dev