[GRASSLIST:3819] Re: Projection lat/lon (gps) to local system

Paul Kelly paul-grass at stjohnspoint.co.uk
Tue Jul 6 19:30:16 EDT 2004


Hello Morten

Morten Sickel wrote:
> 
[...]
> I need to project some gps data to use them together with some raster maps in
> a local (Norwegian) projection :
> 
> GRASS 5.7.cvs:~/NinaSather/Kart >g.proj -p
> -PROJ_INFO-------------------------------------------------
> name       : Transverse Mercator
> datum      : ngo48
> towgs84    : 278.3,93.0,474.5,7.889,0.050,-6.610,6.21
> proj       : tmerc
> ellps      : modif_bessel
> a          : 6377492.0175999999
> es         : 0.0066743722
> f          : 299.1528128500
> lat_0      : 58.0000000000
> lon_0      : 8.3895830000
> k_0        : 1.0000000000
> x_0        : 0.0000000000
> y_0        : 0.0000000000
> 
> As fas as I can see, those values are correct, but anyhow, the gps data ends
> up about 300 meters off from where they should be :-( (see

Please describe how you re-projected your GPS points into the ngo48
datum so we can spot if there is anything wrong there---this seems to be
where the problem is.


> http://sickel.net/gps.html for an illustration) It looks like some kind of
> datum / ellipse error, but I have no idea how to fix that (Except for
> identifying a few pass points and write a transformation script) I have a
> table of dx,dy,.dz values for my local projection and I have a feeling that
> those are somehow connected to the towgs84 parameter, but I don't see how to
> set those.

Edit the PROJ_INFO file, and either change it to have a line
towgs84: dx,dy,dz
or
dx: 
dy:
dz:
with the 3 numbers.

But don't do this until you get the datum transformation working because
there is obviously some problem there. Once you have this done you can
compare your values with the general ones for Norway g.setproj has put
in and see which one makes your data line up better.

Paul




More information about the grass-user mailing list