[Proj] Creating a nadgrid shift file

[Jan Hartmann]

Perhaps I'm doing something stupid, perhaps it cannot be done at all 
with grid shift files. I would very much appreciate if you would have a 
look at it.

I want to rubbersheet old maps with source and destination control 
points in the same projection (epsg:28992). A test example is at

http://mapserver.sara.nl/ntv2.png

The green points are the original points, the blue points are the 
locations I want them shifted to. The shift values are in arcseconds, 
located at the WGS84 coordinates of the four original points

id |  lon    |    lat      |     dx         | dy
--+---------+-----------+-------------+--------------
  ll | 5.3876 | 52.1561 |  -52.6024 | -32.3636
  lr | 5.4022 | 52.1561 |  52.5918 | -32.3701
ur | 5.4022 | 52.1651 |  52.6448 |  32.3504
ul | 5.3876 | 52.1651 | -52.63421 |  32.3561

I created a regular grid with gdal_grid, covering both source and 
destination points, and with these four points as seeds. Then I 
converted that grid to ntv2 format, both with gdal and Jose's program. 
The grid and ntv2 files have correct values at the correct locations.

I would expect that a command like the following:

cs2cs +proj=sterea +lat_0=52.15616055555555 +lon_0=5.38763888888889 
+k=0.9999079 +x_0=155000 +y_0=463000 +ellps=bessel +units=m +no_defs 
+nadgrids=@null \
     +to +proj=sterea +lat_0=52.15616055555555 +lon_0=5.38763888888889 
+k=0.9999079 +x_0=155000 +y_0=463000 +ellps=bessel +units=m +no_defs 
+nadgrids=test_shift.ntv2

would convert source coordinates to destination, but no way. The best I 
get are the red points.

Is the idea correct? If so, how do I get it to work?

Thanks in advance,

Jan



On 01/18/2013 06:01 PM, Jan Hartmann wrote:
> Hi Jose, Gabor, Andre,
>
> I got Jose's program running without trouble, and the results are the 
> same as what I got with gdal. So the problem is not with gdal, but 
> there is something wrong with my data. I'll make a detailed use case 
> tomorrow and hope you can have a look at it.
>
> Jan
>
> On 01/18/2013 11:51 AM, Jose Gonçalves wrote:
>> Hello
>>
>> I have a program to create GSB files from two long/lat difference 
>> grids in Surfer format.
>> http://www.fc.up.pt/pessoas/jagoncal/coordenadas/write_gsb.exe
>>
>> You can check with it. The longitude difference in the input grid are 
>> in the standard way (east positive).
>>
>>
>> Regards
>>
>> José A. Gonçalves
>>
>>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20130119/32002158/attachment.html>