[mapserver-users] RE: RE : [OSRS-PROJ] French NTF projections translation from EPSG to PROJ.4

[Marc Jacquin]

Thanks Claude and Frank,

Claude, I found your definitions in a previous mail somewhere in the
OSRS-PROJ mailing list archive. I send a copy to the mapserver mailing list.
Frank, could it be integrated in some future version of PROJ.4 ?

Regards,

Marc


-----Original Message-----
From: Claude Philipona [mailto:claude.philipona at camptocamp.com]
Sent: jeudi 25 juillet 2002 17:26
To: osrs-proj at remotesensing.org
Cc: 'Marc Jacquin'
Subject: RE : [OSRS-PROJ] French NTF projections translation from EPSG
to PROJ.4


I use these definition for French NTF Projection and it is totally ok
(for me at least). I use them together with IGN scan25 map (2.5m /pixel)
oter data comming in wgs84 or in UTM32N (wgs84) and everything overlaps
good in mapserver. I made some little tests and I can say that the error
is at least better than 2.5 meter for sure, but it seem to be under 1m
most of the time.

Here are the definitions I use (I hacked the epsg file as following), if
it can help:
# NTF (Paris) / France I
<27581> +proj=lcc +a=6378249.2 +b=6356515.0 +lat_0=49d30'0.0"N
+lon_0=2d20'14.025"E +lat_1=48d35'54.682"N +lat_2=50d23'45.282"N
+x_0=600000.0 +y_0=200000.0 +towgs84=-168,-60,+320 +units=m no_defs <>
# NTF (Paris) / France II
<27582> +proj=lcc +a=6378249.2 +b=6356515.0 +lat_0=46d48'0.0"N
+lon_0=2d20'14.025"E +lat_1=45d53'56.108"N +lat_2=47d41'45.652"N
+x_0=600000.0 +y_0=200000.0 +towgs84=-168,-60,+320 +units=m no_defs <>
# NTF (Paris) / France III
<27583> +proj=lcc +a=6378249.2 +b=6356515.0 +lat_0=44d06'0.0"N
+lon_0=2d20'14.025"E +lat_1=43d11'57.449"N +lat_2=44d59'45.938"N
+x_0=600000.0 +y_0=200000.0 +towgs84=-168,-60,+320 +units=m no_defs <>
# NTF (Paris) / France IV
<27584> +proj=lcc +a=6378249.2 +b=6356515.0 +lat_0=42d09'54.0"N
+lon_0=2d20'14.025"E +lat_1=41d33'37.396"N +lat_2=42d46'03.588"N
+x_0=234.358 +y_0=185861.369 +towgs84=-168,-60,+320 +units=m no_defs <>
# NTF (Paris) / France II etendu
<27585> +proj=lcc +a=6378249.2 +b=6356515.0 +lat_0=46d48'0.0"N
+lon_0=2d20'14.025"E +lat_1=45d53'56.108"N +lat_2=47d41'45.652"N
+x_0=600000.0 +y_0=2200000.0 +towgs84=-168,-60,+320 +units=m no_defs <>

I also tried to use them with data coming from ch1903 Oblique
cylindrical (swiss projection) with satisfaction:

# CH 1903 / Swiss Oblique Cylindrical
<9814> +proj=somerc +lat_0=46d57'08.66"N +lon_0=7d26'22.50"E
+ellps=bessel +x_0=600000 +y_0=200000 +towgs84=674.374,15.056,405.346
+units=m +k_0=1 +no_defs <>

Claude Philipona

> > I can't find out why in the EPSG definition file French projections
> > (NTF
> > (Paris) / France I to IV) cannot be translated into PROJ.4.
> In the EPSG file
> > it's commented as "Unable to to translate coordinate system
> into PROJ.4
> > format". Obviously the NTF geodetic system is recognized
> and Lambert Conform
> > Conic too. Any help on this matter would surely be
> appreciated by French
> > mapserver users.
>
> Marc,
>
> Generally this happens when there is more than one
> approximation available for a datum (to WGS84) in the EPSG
> tables, as the auto-translation code can't decide which to
> use.  I also doubt that the EPSG to PROJ.4 stuff properly
> accounts for non-greenwich prime meridians.