[Proj] A bit of help with French, por favor.

[Gerald I. Evenden]

On Friday 20 June 2008 1:41 pm, OvV_HN wrote:
> This has been discussed earlier in the case of the French  Laborde
> projections.
> (I have currently no access to my files or emails, so from memory...)
> If I remember correctly, the libproj stmerc projection was the same as one
> of the 3 French (IGN) Laborde projections.

Yes, I dug out that old allegation also but I am somewhat skeptical.  At the 
moment I cannot get stmerc to replicate the example numbers in NTG-73.

By using the proj_gauss function the double projection is like falling off a 
log and the code makes it obvious that stmerc is using the spherical 
transverse Mercator formulas.  But the NTG-73 procedure is of the old school 
and the formulas are intertwined and it is unclear the nature of the final 
spherical formula.

Unlike the transverse Mercators we are used where h=k=1 along the CM, stmerc 
has only one point at the projection origin where h=k=1.

I am also doubtful that Schreiber is the correct designation.

> I assume that these projections are double projections and work with an
> intermediate sphere.
> The sphère de courbure has an intermediate sphere with a radius equal to
> the radius of curvature at the touching point - or something like that. The
> sphère équatorial has a radius equal to the semi-major (equatorial) axis of
> the ellipsoid.
> The sphère bitangente touches the ellipsoid twice, I guess. Don't ask me
> where, but this could give a clue looking at the formulae.
> See:
> http://www.ign.fr/rubrique.asp?rbr_id=1700&lng_id=FR#68096
> and NT/G 73 for the French Laborde projection formulae.
> Hope this will help.
	...
-- 
The whole religious complexion of the modern world is due
to the absence from Jerusalem of a lunatic asylum.
-- Havelock Ellis (1859-1939)  British psychologist