[Proj] Unknown projection?

Gerald I. Evenden geraldi.evenden at gmail.com
Sat Sep 13 10:40:12 PDT 2008


On Saturday 13 September 2008 12:26:58 pm Paul B. Anderson wrote:
> - Janne,
>
> The difficulty comes about because the Tripel is actually composed of 2 map
> projections -- the Aitoff (pronounced 'I-toff') projection averaged with
> the Equirectangular projection (which provides the Special Parallel input).
> Yes, there are inverse formulas out there for the Tripel, but the values
> produced are approximations.
>
> One person who has researched the inverse Winkel Tripel formulas issue is
> Dr. Robert Schmunk http://www.giss.nasa.gov/staff/rschmunk.html
>
> BTW, a reference for this map projection is Snyder's 'Flattening the
> Earth'.
>
> Paul B. Anderson FCCS (USN, Retired)
> Kingsport, TN native living in Norfolk, VA
> My Map Projection web site:
> http://www.csiss.org/map-projections/index.html
>
> ----- Original Message -----
> From: <support.mn at elisanet.fi>
> To: "PROJ.4 and general Projections Discussions" <proj at lists.maptools.org>;
> <geraldi.evenden at gmail.com>
> Sent: Saturday, September 13, 2008 11:42 AM
> Subject: Re: [Proj] Unknown projection?
>
> > "Gerald I. Evenden" <geraldi.evenden at gmail.com> kirjoitti:
> >> On Saturday 13 September 2008 10:59:24 am Paul B. Anderson wrote:
> >> > - Janne,
> >> >
> >> > Your map graphic appears to be in the Winkel Tripel map projection. I
> >> > would
> >> > also suggest that it is using Oswald Winkel's original selection of
> >> > 50.46667 degrees as the Special Parallel.
> >> >
> >> > Note that by the 1950s John Bartholomew was using a special parallel
> >> > of 40
> >> > degrees for his version of the Tripel, giving his graphic a more
> >> > flattened
> >> > appearance in the Y direction.
> >>
> >> This is exactly why I include the projection math in the manual.  The
> >> factor Anderson is discussing is +lat_1 option for the +proj=wintri
> >> projection.  Select your own option value and name a new projection. 
> >> ;-)
> >
> > Ok, now the problem seems to be that "wintri" does not have any inverse
> > projection
> > available. What is the reason not to have inverse projections for all
> > routines
> > in the library... math is too hard?
> >
> > Janne.

I can't find any reference on his website to research on Winkel Tripel so I 
suspect he has not published it.

I take back my comment on ease of solution but it would seem to be invertible 
using iterative techniques---but slow.  Accuracy should only be dependent 
upon how many cpu ticks you want to give it.

-- 
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



More information about the Proj mailing list