[Proj] Amersfoort / RD New

[Bas Couwenberg]

On 2016-11-01 15:05, Roger Oberholtzer wrote:
> I downloaded the grid shift file from
> http://www.kadaster.nl/transformatie-van-coordinaten
> 
> I am not using it yet. I took one of the sample points from the Use doc
> that came with it to see what I got compared to the expected. For 
> example:
> 
> Zuid-Limburg:
> Ref:   Latitude: 50.7925849160 Longitude: 5.7737955480 Altitude: 
> 245.948
> Ref:   Easting:     182260.450 Northing:    311480.670 Altitude: 
> 200.000
> Proj4: Easting:     182225.909 Northing:    311388.163 Altitude: 
> 245.948
> 98.746 meters from Ref
> 
> The reference values are ETRS89. But I cannot think that accounts to 
> 100
> meters of difference. My proj spec is:
> 
>      +proj=sterea
>      +lat_0=52.15616055555555
>      +lon_0=5.38763888888889
>      +k=0.9999079
>      +x_0=155000
>      +y_0=463000
>      +ellps=bessel
> 
> +towgs84=565.4174,50.3319,465.5542,-0.398957388243134,0.343987817378283,-1.87740163998045,4.0725
>      +units=m
>      +no_defs
> 
> Doing this via the C API as I do for many other projections. These are
> passed to pj_init_plus.
> 
> Would the grid shift make such a large correction?

 From the proj-rdnap test script (using PROJ.4 4.9.3):

  Test:   05 Zuid-Limburg
  Exec:   cs2cs -r +init=epsg:4258 +to +init=rdnap:rdnap -f %.4f
  Input:  50.792584908 5.773795547 174.9478
  Output: 182260.4500     311480.6701 129.0000
  Expect: 182260.4500     311480.6700 129.000
  Test OK: From ETRS89 to RD/NAP - 05 Zuid-Limburg (Not identical, but 
within margin)

The above uses the test values from "Use of RDTRANS2008 and 
NAPTRANS2008.pdf" included in the NTv2.zip, your values are different, 
not sure where you got those.

The rdnap configuration for the above test contains:

  # RDNAP with NTv2 and VDatum
  <rdnap> +proj=sterea +lat_0=52.15616055555555 +lon_0=5.38763888888889 
+k=0.9999079 +x_0=155000 +y_0=463000 +ellps=bessel 
+nadgrids=rdtrans2008.gsb +geoidgrids=naptrans2008.gtx +units=m +no_defs 
<>

Try using the grid shift files for your test, it should be more accurate 
(at ground level) than using +towgs84 values.

Kind Regards,

Bas