EPSG code 3785 not translated in "epsg" file about proj

[proj4-bot]

Comment by neteler on 2 Aug 2008 15:23 UTC
EPSG recently published Version 6.17 - perhaps it could go into the new PROJ release in case above problem requires a re-run of the nad/epsg file creation?

Best
Markus

[proj4-bot]

Comment by warmerdam on 14 Aug 2008 14:40 UTC
The problem seems to be that this projection uses projection method code 9841 which is not currently supported in ogr_fromepsg.cpp (SetEPSGProjCS()). I'll do some investigation.

[proj4-bot]

Comment by warmerdam on 14 Aug 2008 18:48 UTC
The SetEPSGProjCS() function in OGR has been altered in http://trac.osgeo.org/gdal/changeset/15148 to treat 9841 (spherical mercator 1SP) the same as regular mercator 1SP. However, this results in the following definition:

PROJCS[Visualisation CRS / Mercator",
    GEOGCS["Popular Visualisation CRS",
        DATUM["Popular_Visualisation_Datum",
            SPHEROID["Popular Visualisation Sphere",6378137,0,
                AUTHORITY["EPSG","7059"]("Popular)],
            TOWGS84[           AUTHORITY["EPSG","6055"](0,0,0,0,0,0,0],
)],
        PRIMEM[           AUTHORITY["EPSG","8901"]("Greenwich",0,
)],
        UNIT[           AUTHORITY["EPSG","9122"]("degree",0.01745329251994328,
)],
        AUTHORITY[   UNIT["metre",1,
        AUTHORITY["EPSG","9001"]("EPSG","4055"]],
)],
    PROJECTION[   PARAMETER["central_meridian",0]("Mercator_1SP"],
),
    PARAMETER[   PARAMETER["false_easting",0]("scale_factor",1],
),
    PARAMETER[   AUTHORITY["EPSG","3785"]("false_northing",0],
),
    AXIS[   AXIS["Y",NORTH]("X",EAST],
)]

The problem here is that a TOWGS84[] entry is introduced. And this means that conversion to WGS84 will result in conversion from lat/long on th sphere to geocentric coordinates (using the sphere), application of a 0,0,0 offset, and then conversion back to lat/long on the wgs84 ellipsoid which will be a substantialy offset from the original latitude.

The intention is that the spherical coordinates be treated as if they were on WGS84.

I'm presuming the TOWGS84 node is introduced because this conversion existed in the EPSG dictionary, but it seems wrong to me. I'm going to ask for advice on the metacrs list.

[proj4-bot]

Comment by warmerdam on 7 Aug 2011 20:57 UTC
The coordinate system is supported properly in PROJ.4 now, and is present in the epsg init file. I'm still not super thrilled with the way that it is represented in WKT, but that isn't a PROJ.4 problem.

I'm not positive if it was in the 4.7 init file, but it will definately be in the 4.8 init file.