[PROJ] Help needed for a cs2cs pipeline with shift grids

Tue Feb 20 09:53:04 PST 2024

Hi Even 

Thanks for your kind and developed answer. 
I've tried with the EPSG equivalents: ( [ https://www.sandre.eaufrance.fr/jeu-de-donnees/projection-des-coordonn%C3%A9es?lang=fr | https://www.sandre.eaufrance.fr/jeu-de-donnees/projection-des-coordonnées?lang=fr ] ) 

echo 0 56.000000 0 | ./cs2cs +from EPSG:4807 +to EPSG:3034 

But the result doesn't make any sense: 

-1640975.39 10099142.17 0.00 

Is there something with the units that is not correct maybe? 

This works well: 

echo 0 56.000000 0 | PROJ_NETWORK=ON ./cs2cs -d 3 "NTF geographiques Paris (gr) + NGF-IGN69 height" IGNF:ETRS89LCC --3d 

and I have a few questions about it: 

1) the result remains unchanged if I remove or put PROJ_NETWORK to OFF. What does it do exactly? 

2) Does the string " NTF geographiques Paris (gr) + NGF-IGN69 height" correspond to an entry in the CRS database proj.db or is it interpreted as a compounded system because PROJ parses the string? There is no simple code like EPSG or IGNF related to it? Can we create one easily? 

3) Where can I see where this crs"NTF geographiques Paris (gr) + NGF-IGN69 height" is defined in the sql commands used to build it? 

Nicolas Bellaiche 

De: "Even Rouault" <even.rouault at spatialys.com> 
À: "Nicolas Bellaiche" <nicolas.bellaiche at ign.fr>, "proj" <PROJ at lists.osgeo.org> 
Envoyé: Dimanche 18 Février 2024 15:16:44 
Objet: Re: [PROJ] Help needed for a cs2cs pipeline with shift grids 

Nicolas, 

I believe the Helmert transformation is used when using IGNF codes, because the IGNF database has only recorded the use of the ntf_r93 grid for NTF to RGF93, but not for NTF to ETRS89 (by the way, just reiterating that IGN is more than welcome to take over on maintenance & updates of the IGNF part of the PROJ database). And when transforming between 2 IGNF codes, PROJ by default will only consider transformations in the IGNF domain. But if you use EPSG codes, then ntf_r93 is available to transforme between NTF and ETRS89. 

To get a vertical transformation, you also need to use 3D CRS. 

For example 

projinfo -s "NTF geographiques Paris (gr) + NGF-IGN69 height" -t IGNF:ETRS89LCC --3d --spatial-test intersects 

reports the following pipeline: 

unknown id, axis order change (2D) + NTF (Paris) to NTF (1) + NTF to RGF93 v1 (1) + Inverse of RGF93 v1 to NGF-IGN69 height (1) + RGF93 v1 to ETRS89 (1) + axis order change (geographic3D horizontal) + ETRS89 LAMBERT CONFORMAL CONIC, 1.6 m, France - mainland onshore., at least one grid missing 

PROJ string: 
+proj=pipeline 
+step +proj=unitconvert +xy_in=grad +xy_out=rad 
+step +inv +proj=longlat +ellps=clrk80ign +pm=paris 
+step +proj=push +v_3 
+step +proj=cart +ellps=clrk80ign 
+step +proj=xyzgridshift +grids=fr_ign_gr3df97a.tif +grid_ref=output_crs 
+ellps=GRS80 
+step +inv +proj=cart +ellps=GRS80 
+step +proj=pop +v_3 
+step +proj=vgridshift +grids=fr_ign_RAF18.tif +multiplier=1 
+step +proj=lcc +lat_0=52 +lon_0=10 +lat_1=35 +lat_2=65 +x_0=4000000 
+y_0=2800000 +ellps=GRS80 

Here, as there's a mix of objects from different authorities (IGNF and EPSG), then EPSG transformations are considered, hence you get the gr3df97a grid. 

$ echo 0 56.000000 0 | PROJ_NETWORK=ON bin/cs2cs -d 3 "NTF geographiques Paris (gr) + NGF-IGN69 height" IGNF:ETRS89LCC --3d 
3474789.997 2655359.291 43.644 

which is the application of the above pipeline: 

$ echo 0 56.000000 0 | PROJ_NETWORK=ON bin/cct -d 3 +proj=pipeline +step +proj=unitconvert +xy_in=grad +xy_out=rad +step +inv +proj=longlat +ellps=clrk80ign +pm=paris +step +proj=push +v_3 +step +proj=cart +ellps=clrk80ign +step +proj=xyzgridshift +grids=fr_ign_gr3df97a.tif +grid_ref=output_crs +ellps=GRS80 +step +inv +proj=cart +ellps=GRS80 +step +proj=pop +v_3 +step +proj=vgridshift +grids=fr_ign_RAF18.tif +multiplier=1 +step +proj=lcc +lat_0=52 +lon_0=10 +lat_1=35 +lat_2=65 +x_0=4000000 +y_0=2800000 +ellps=GRS80 

If you want to get exactly the below pipeline, you'll have to tweak proj.db to create a compound CRS for "NTF geographiques Paris (gr) + NGF-IGN69 height", and then create a custom record in other_transformation for the transformation between this compound CRS and (a 3D projected CRS derived from) IGNF:ETRS89LCC. Cf data/sql/other_transformation_custom.sql for potential inspiration 

Even 

Le 15/02/2024 à 19:46, Nicolas Bellaiche via PROJ a écrit : 

Hi there, 

I try to make a conversion using shiftgrids from IGNF: NTFP to IGNF:ETRS89LCC. By default, it seems that it uses the helmert approximation between the 2 crs and i cannot figure out how to use the grid fr_ign_ntf_r93. 

So far i've found how to create a pipeline that works with cct, but i'd like to have a code that represents the crs and be able to go to different destination crs with the cs2cs application. Any idea? 

(helmert approx + no geoid with the standard codes) 

echo 0 56 0 | ./cs2cs + from IGNF:NTFP +to IGNF:ETRS89LCC 

result : 3474788.21 2655358.23 0.00 

(correct pipeline) 

echo 0 56.000000 0 | PROJ_DEBUG=3 ./cct +proj=pipeline +step +proj=unitconvert +xy_in=grad +xy_out=rad +step +inv +proj=longlat +ellps=clrk80ign +pm=paris +step +proj=hgridshift +grids=fr_ign_ntf_r93.tif +step +inv +proj=vgridshift +grids=RAF09.gtx +step +proj=lcc +lat_0=52 +lon_0=10 +lat_1=35 +lat_2=65 +x_0=4000000 +y_0=2800000 +ellps=GRS80 

result : 3474789.9972 2655359.2908 43.6421 inf 

reference from THE official IGN/SGN dataset: 

3474789.997 2655359.291 43.642 

Thanks a lot for your help, 

Nicolas Bellaiche 

_______________________________________________
PROJ mailing list [ mailto:PROJ at lists.osgeo.org | PROJ at lists.osgeo.org ] [ https://lists.osgeo.org/mailman/listinfo/proj | https://lists.osgeo.org/mailman/listinfo/proj ] 

-- [ http://www.spatialys.com/ | http://www.spatialys.com ] My software is free, but my time generally not. 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20240220/56e05fbc/attachment-0001.htm>