On the design of globe gores

General discussion of map projections.
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

On the design of globe gores

Post by daan »

This is the slide deck for a presentation I gave at NACIS 2001. While the screen shots are outdated, and of course Geocart 3 is considerably more capable, the principles themselves have not changed.

Download is 4.5 MB.

Happy mapping!
— daan
gearmar
Posts: 2
Joined: Thu Jan 02, 2020 5:33 pm

Re: On the design of globe gores

Post by gearmar »

I would be very interested in reading through the slide deck. The hyperlink is broken. Can you please re-attach the link?

Thanks, Mark
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: On the design of globe gores

Post by daan »

Well hrmph. I am not sure how that got lost. I have updated the link in my original post. Enjoy!

— daan
gearmar
Posts: 2
Joined: Thu Jan 02, 2020 5:33 pm

Re: On the design of globe gores

Post by gearmar »

Thanks very much. I see it now.
Piotr
Posts: 313
Joined: Thu Mar 23, 2017 12:27 pm

Re: On the design of globe gores

Post by Piotr »

globe gores under 1 parameter is American Polyconic, while as the parameter approaches Infinity, it becomes Sinusoidal. In between are compromise projections.
Piotr
Posts: 313
Joined: Thu Mar 23, 2017 12:27 pm

Re: On the design of globe gores

Post by Piotr »

If the parameter is below 1, globe gores is self-intersecting (but still bijective, as it doesn't fail to render). So if the paper is so stiff that its warpage is below 1, then you have to print the map in multiple pieces and put it together to make the self-intersecting globe gores.
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: On the design of globe gores

Post by daan »

Piotr wrote: Tue Mar 31, 2020 6:32 am If the parameter is below 1
BUG.

Fixed.
globe gores is self-intersecting (but still bijective, as it doesn't fail to render)
Geocart does not perform some mathematical check for bijective property. It simply tries to render. It may not find a continuous boundary, in which case it will not render at all. It might find a continuous, closed boundary, but the boundary might intersect itself, which ought to be disallowed. However, Geocard does not try to detect self-intersection. In some of those cases, it will draw; in others it will not. If the boundary is continuous and not self-intersecting, it will draw, but the interior could be quite messed up, depending on all kinds of things.

— daan
Piotr
Posts: 313
Joined: Thu Mar 23, 2017 12:27 pm

Re: On the design of globe gores

Post by Piotr »

daan wrote: Tue Mar 31, 2020 8:25 pm
Piotr wrote: Tue Mar 31, 2020 6:32 am If the parameter is below 1
BUG.

Fixed.
But what if the paper is so stiff that its warpage is below 1, then does that mean users would have to downgrade Geocart or edit the source code themselves (if it ever becomes open-source) to do this?
daan wrote: Tue Mar 31, 2020 8:25 pm
globe gores is self-intersecting (but still bijective, as it doesn't fail to render)
Geocart does not perform some mathematical check for bijective property. It simply tries to render. It may not find a continuous boundary, in which case it will not render at all. It might find a continuous, closed boundary, but the boundary might intersect itself, which ought to be disallowed. However, Geocard does not try to detect self-intersection. In some of those cases, it will draw; in others it will not. If the boundary is continuous and not self-intersecting, it will draw, but the interior could be quite messed up, depending on all kinds of things.

— daan
What's the problem with self-intersecting boundaries? If you were to extrapolate the sinusoidal projection away from the plane chart, you get a self-intersecting boundary as well, but the actual area doesn't overlap. Is Geocard a mobile version of Geocart?
Piotr
Posts: 313
Joined: Thu Mar 23, 2017 12:27 pm

Re: On the design of globe gores

Post by Piotr »

Make globe gores support complex numbers, positive i warpage is that the paper warpage is more conformal, while negative i warpage is that the paper warpage is more equal-area.
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: On the design of globe gores

Post by daan »

Piotr wrote: Tue Mar 31, 2020 9:50 pm What's the problem with self-intersecting boundaries?
It implies that the range of the map overlaps itself, which is no longer bijective. (The boundary is allowed not to be bijective and is allowed to touch itself. It must not cross itself.)
If you were to extrapolate the sinusoidal projection away from the plane chart, you get a self-intersecting boundary as well, but the actual area doesn't overlap.
I don’t follow “away from the plane chart”.

— daan
Post Reply