[gdal-dev] Re: Slovenia vector data reprojection issue

[Frank Warmerdam]

Markus Neteler wrote:
> Hi,
> 
> (back to an old problem)
> 
> On Fri, Jul 24, 2009 at 4:31 PM, Markus Neteler <neteler at osgeo.org> wrote:
>> Hi,
>>
>> I am trying to reproject Slovenian country data from D48 (WKT attached)
>> and Italian ISTAT national vector data (UTM32/ED50) into a common
>> projection.
>>
>> While the map "attach" rather well in Latlong/WGS84 using GDAL 1.6.svn,
>> they don't in EU LAEA (EPSG 3035). This is a bit suprising for me.
>> Any trick to get this done?
>>
>> The input SHAPE files have each a .prj file (attached), the target SRS
>> I request by -t_srs epsg:4326 or -t_srs epsg:3035.
>>
>> My preferred SRS would be EU LAEA to match the EU CORINE data.
> 
> It appears to be a bug in GDAL.
> If I do
> 
> * SLO -> EU LAEA
>   the result is shifted, apparently the TOWGS84 in epsgD48_3787.prj is
> not applied.
> 
> If I do instead:
> * SLO -> LL -> EU LAEA
>   the result is "perfect".

Markus,

The problem is that the eu_etrs89_laea.prj does not have TOWGS84
parameters for the datum, so if you go directly from D48 to laea
no datum shifting will be done (since one end is not datum enabled).

If you go to WGS84 first, then the D48 datum shift is applied to WGS84.
No datum shift is applied from WGS84 to the LAEA projection, but ETRS89
is effectively the same as WGS84 I believe.

So the key is to mark the ETRS89 datum as equal to WGS84 - likely with a
TOWGS84[0,0,0] entry in the coordinate system definition.

I see that after my datum overhaul, GDAL is expanding EPSG:4258 to:

GEOGCS["ETRS89",
     DATUM["European_Terrestrial_Reference_System_1989",
         SPHEROID["GRS 1980",6378137,298.257222101,
             AUTHORITY["EPSG","7019"]],
         TOWGS84[0,0,0,0,0,0,0],
         AUTHORITY["EPSG","6258"]],
     PRIMEM["Greenwich",0,
         AUTHORITY["EPSG","8901"]],
     UNIT["degree",0.01745329251994328,
         AUTHORITY["EPSG","9122"]],
     AUTHORITY["EPSG","4258"]]

which has the TOWGS84[] entry, so we can have confidence that this
family of problem will be reduced as the richer set of datum shifts
comes out with GDAL 1.8.

In short, I think this is an understood situation and there is no
need for a ticket.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent