On 10/06/09 05:02, Hamish wrote:
Paul wrote:
Probably we should update datumtransform.table with the
correct Belgium 72 parameters. But it would be reassuring to
find out where the wrong ones have come from first...
.... and which ones are in fact "wrong" ...
I'm a bit surprised that Belgium is big enough to have two
regional sets with significant differences.
Here's the info that the German site got from the Belgian agency when
they asked a question about this a few years ago:
"Both of the sets of parameters can be used. The result is the same and
the quality is about 20 cm (max. error). For more precise applications
an extra step has to be done after the helmert transformation: pass the
results through a correction grid (bilinear interpollation) to fit them
better to LB72. This should be done with the set of parameters indicated
on our website; in that case the accuracy of transformation is 1.2 cm
(stdev)."
But when I test with cs2cs, here are the results:
- echo "4.468493 50.856057" | cs2cs +init=epsg:4326 +to +init=epsg:31370
157201.51 171803.24
- (checking to see if there is any difference with ETRS89 instead of WGS84):
echo "4.468493 50.856057" | cs2cs +init=epsg:4258 +towgs84=0,0,0,0,0,0,0
+to +init=epsg:31370
157201.51 171803.24
- Using the official software of the Belgian agency (which uses ETRS89
which is considered +/- equivalent to WGS84):
157022.870 171745.084
- Replacing epsg:31370 with the parameters for that code in
/usr/share/proj/epsg (just to make sure):
echo "4.468493 50.856057" | cs2cs +init=epsg:4326 +to +proj=lcc
+lat_1=51.16666723333333 +lat_2=49.8333339 +lat_0=90
+lon_0=4.367486666666666 +x_0=150000.013 +y_0=5400088.438 +ellps=intl
+towgs84=106.869,-52.2978,103.724,-0.33657,0.456955,-1.84218,1 +units=m
+no_defs
157201.51 171803.24 (as expected)
- Changing these parameters to reflect the exact parameters given on the
Belgian agency's website (http://www.ign.be/FR/FR2-1-4.shtm and
http://www.ign.be/FR/FR2-1-2.shtm):
echo "4.468493 50.856057" | cs2cs +init=epsg:4326 +to +proj=lcc
+lat_2=51.16666723333333 +lat_1=49.8333339 +lat_0=90
+lon_0=4.367486666666666 +x_0=150000.013 +y_0=5400088.438 +ellps=intl
+towgs84=106.869,-52.2978,103.724,0.33657,-0.456955,1.84218,1 +units=m
+no_defs
157143.81 171773.43
So we're talking about serious differences here...
- Applying the parameters from
http://crs.bkg.bund.de/crseu/crs/eu-countrysel.php?country=BE#StartCE:
echo "4.468493 50.856057" | cs2cs +init=epsg:4326 +to +proj=lcc
+lat_1=49.8333339 +lat_2=51.16666723333333 +lat_0=50,799362222
+lon_0=4.367486666666666 +x_0=150000.013 +y_0=165372.95628 +ellps=intl
+towgs84=-99.1,53.3,-112.5,0.419,-0.830,1.885,-1 +units=m +no_defs
157022.95 260661.69
So, completely different, and way beyond the 20cm mentioned above that
you can correct with the correction grid.
Me completely confused...
But maybe I should take this to the proj list and/or discuss it with our
local agency's experts...
But to come back to my GRASS-related question: do I understand correctly
that datumtransform.table is used only once during location creation ?
Moritz
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev
