[Proj] Differences between UTM Zone 30 (ED 50) projections

[Jelmer Baas]

Hello Andre,

Thanks for the reply. I'll study your replies some more tomorrow, but I also continued experimenting, and found that the following yields nearly exactly the same results as MapInfo:
cs2cs +nadgrids=NTV2_0.GSB,BALEARES.gsb,PENINSULA.gsb +proj=utm +zone=30 +ellps=intl +units=m +to +proj=latlong +ellps=WGS84 +towgs84=0,0,0
[.]
pj_open_lib(BALEARES.gsb): call fopen(C:\OSGeo4W64\share\proj\BALEARES.gsb) - succeeded
NTv2 BALEARES 93x68: LL=(0.833333333,38) UR=(4.66666667,40.7916667)
pj_open_lib(PENINSULA.gsb): call fopen(C:\OSGeo4W64\share\proj\PENINSULA.gsb) - succeeded
NTv2 PENINSUL 259x161: LL=(-10.1833333,35.5555556) UR=(4.15,44.4444444)
NTv2 - loading grid PENINSUL
pj_open_lib(PENINSULA.gsb): call fopen(C:\OSGeo4W64\share\proj\PENINSULA.gsb) - succeeded
pj_apply_gridshift(): used PENINSUL
3d42'30.984"W   40d35'42.94"N 0.000

cs2cs +proj=latlong +ellps=WGS84 +towgs84=0,0,0 +to +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0 +y_0=0 +k=1.0 +units=m
3d42'30.984"W   40d35'42.94"N 0.000
-412840.21      4952824.83 0.00

So when doing UTM Zone 30 ED 50 -> WGS84 -> Mercator, I get results I expect. Note that I'm using the .gsb files provided with MapInfo PRO, which answers you question below: the projection system as defined in MapInfo doesn't have the datum shift, but it uses the shift defined in its .gsb files.
Can I force this extra step in a proj wkt string? My goal is to have this working in PostgreSQL.

Using a fixed shift value could also work, since I'm only intested in Spain at the moment.

I'm not sure I understand this question of yours: " What datum shift grid did you use? the files I know are mainly from national coordinate systems to WGS84.". Care to eloborate?

Thanks,

Regards,
Jelmer

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Andre Joost
Sent: donderdag 13 oktober 2016 20:10
To: proj at lists.maptools.org
Subject: Re: [Proj] Differences between UTM Zone 30 (ED 50) projections

Am 13.10.2016 um 10:07 schrieb Jelmer Baas:
> Hello,
>
> I currently have a MapInfo file in UTM Zone 30 (ED 50) (EPSG: 23030)
> projection, which, after reprojection via OGR2OGR to EPSG:900913,
> moves a few meters East when compared to OSM.

Looks like a datum shift issue, but kkep in mind that OSM is not 
necessarily "exact".


>
> Oddly enough, the EPSG file defines 23030 as +proj=utm +zone=30
> +ellps=intl +towgs84=-87,-98,-121,0,0,0,0 +units=m +no_defs  <>

... using a mean shift value for whole Europe.

>
> Yet, spatialreference.org
> (http://spatialreference.org/ref/epsg/ed50-utm-zone-30n/ ) defines it
> as +proj=utm +zone=30 +ellps=intl +units=m +no_defs

... omitting the datum shift, which is surely wrong.

>
> For completeness, MapInfo PRO defines this projection system as: UTM
> Zone 30 (ED 50)\p23030", 8, 28, 7, -3, 0, 0.9996, 500000, 0

I'm not sure if this includes an explicit datum shift.


> From a developer at MapInfo I got a tip to try to include an accurate
> datum transform file, like NTV2_0.GSB, but CS2CS doesn't want to use
> this: This file doesn't even exist: cs2cs +nadgrids=blah.gsb
> +proj=utm +zone=30 +ellps=intl +towgs84=-87,-98,-121,0,0,0,0 +units=m
> +to +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0
> +y_0=0 +k=1.0 +units=m cs2cs +proj=utm +zone=30 +ellps=intl
> +towgs84=-87,-98,-121,0,0,0,0 +units=m +to +proj=merc +a=6378137
> +b=6378137 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0 +y_0=0 +k=1.0 +units=m
> +nadgrids=blah.gsb [no error, Process Monitor shows cs2cs not even
> trying to open the file]

What datum shift grid did you use? the files I know are mainly from 
national coordinate systems to WGS84.



>  Can anyone tell me why there's a difference between the MapInfo PRO
> projection and the Proj.4 projection? And, hopefully, also a solution
> to correct this? Please note that I don't know whether MapInfo or
> Proj is at fault, I justk now that my data is drawn from MapInfo,
> where it's accurate compared to other data.

The EPSG registry lists 42 valid datum shift parameter sets from ED50 to 
WGS84, each with a different area of use. As written above, the values 
in EPSG:23030 are a mean value, which can not provide centimeter accuracy.

HTH,
André Joost


_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj