[Proj] Graduated equidistant projections for convenient co-ordinate transformations

[Michael Ossipoff]

Sorry to post so much, but this completes what I have to say, and so I won't 
keep on making a nuisance of myself :-)  I've noticed that I left something 
out of my previous posting:

I forgot to add that I realize that the sinusoidal projection gives a 
combination of easy position-measurement and equal area. (Bonne and 
Stabius-Werner are much more inconvenient, due to their curved parallels and 
meridians).

Though the sinusoidal ‘s curved meridians make longitude measurement not 
quite as easy as it is with a cylindrical projection, it’s still easy.

Though, for me, the locations of a data map’s zone boundaries are the map’s 
only important information, and the whole point of the map, I understand 
that many people are also interested in global comparisons and the relative 
areas of zones. The sinusoidal suits both purposes.

The sinusoidal’s  scale variation and departures from conformality aren’t so 
bad when the map is interrupted.

If an uninterrupted world map is desired, with easy position-measurement, 
less distortion than the uninterrupted sinusoidal,  and reasonably equal 
area, I’d suggest the equidistant elliptical projection. When Mollweide has 
a 30 degree graticule, the spacing of its parallels is only a little 
non-uniform. That suggests that the equidistant elliptical wouldn’t have 
enough area inaccuracy to bother anyone. Among the uninterrupted world maps, 
an elliptical projection with straight parallels, in my opinion, gives the 
most realistic-looking and attractive overall portrayal of the Earth.

The price of an uninterrupted, reasonable-areas world map is that it isn’t 
much good for distances, directions and shapes. It’s attractive, but the 
interrupted sinusoidal, the graduated equidistant cylindrical, and the 
Mercator are more useful.

Mike Ossipoff