[Proj] Help with use of swisstopo NTv2 grid shift ?

Sandro Santilli strk at keybit.net
Mon Jan 23 09:47:55 PST 2012


On Mon, Jan 23, 2012 at 05:22:03PM +0000, Paul Kelly wrote:

> No, this is exactly the same as the bounds reported by PROJ.4. Those 
> values are in degree-seconds. You need to divide by 3600 to get them in 
> degrees (and note also that the longitude values are positive west of the 
> Greenwich meridian; you need to divide them by -3600).

Ah, thanks !

> I suspect you may have gone a little too far in "pretending" that you're 
> using WGS84! Are your co-ordinates really in latitude/longitude? The very 
> large numbers shown in the error message suggest that you have projected 
> co-ordinates; see below:

Indeed I was feeding wrong input. Local instead of global.

> I think you need to set up your co-ordinate system definition strings to 
> specify the correct projection and ellipsoid that you are using.

The EPSG definition for the source and target projection are these:

 # CH1903
 <4149> +proj=longlat +ellps=bessel +towgs84=674.4,15.1,405.3,0,0,0,0 +no_defs  <>
 # CH1903+
 <4150> +proj=longlat +ellps=bessel +towgs84=674.374,15.056,405.346,0,0,0,0 +no_defs  <>

So, should I keep the +ellps=bessel in my "fake" ones, like this ?

 <CH1903>  +proj=longlat +ellps=bessel +nadgrids=CHENYX06a.gsb +no_defs
 <CH1903+> +proj=longlat +ellps=bessel +towgs84=0,0,0 +no_defs

It would surely help if I had an official set of points in CH1903 and CH1903+ to
test my conversions against... 

--strk;

  ()   Free GIS & Flash consultant/developer
  /\   http://strk.keybit.net/services.html



More information about the Proj mailing list