[Proj] Finnish KKJ in proj 4.4.9 - Linux vs. Windows

[Roger Oberholtzer]

On Tue, 2006-01-03 at 13:06 -0500, Frank Warmerdam wrote:
> On 1/3/06, Roger Oberholtzer <roger at opq.se> wrote:
> >            On Wed, 2005-12-14 at 15:44 +0100, Roger Oberholtzer wrote:
> >  I have sorted out using proj 4.4.9 to obtain KKJ X/Y values from WGS84
> > LAT/LONG values.
> 
> Roger,
> 
> I tried to reproduce your results, but was unsuccessful.
> 
> I converted the minute location (6012.21644 2452.87822) to
> degrees as (100.203607E 40.881303N), somewhere in Mongolia
> or northern China.  Is that right?  You referred to the columns as
> lat and long, but a latitude of 100N makes no sense, so I assumed
> they were really long and lat, respectively.
> 
> I tried transforming this like:
> 
> cs2cs +proj=latlong +ellps=WGS84 +to +init=esri:2393
> 100.20360733333334 40.881303666666668
> 9357680.26      8013647.87 -76.26
> 
> Well, the results aren't even in vaguely the same ballpark as your
> values.
> 
> So I tried going in reverse to see if that worked any better.
> 
> cs2cs  +init=esri:2393 +to +proj=latlong +datum=WGS84
> 6678774.88 3382686.20
> 58d3'14.396"E   26d48'49.909"N 64.519
> 
> Not really close to anything either.
> 
> Now, I could continue to guess what I am interpreting wrong in
> your instructions, but I think that may be pointless.  Could you please
> try your case with cs2cs, and give an easy example we could type in
> to try and reproduce your results?
> 
> PS. esri:2393 is a tranverse mercator projection.  These
> are genreally not stable far from the central meridian.  In this
> case it would appear you are working 2.5 million meters away
> from the central meridian (since the false easting is 3500000m).
> That is a substantial distance.

The values should have been described as ddmm.mmmmm, not minutes. as I
first specified. And they are from near Helsinki.


--
Roger