What's your favorite map projection?

[BobTheBarbarian]

Calling all geography nerds!

Which map projection do you like the best? Either for its geometric properties or for aesthetic appeal? Why is it your favorite?

Which would you use as a reference map to teach geography to others?

I like Van der Grinten I:

Pros: decent at maintaining shape of countries, less distortion than Mercator, was the official projection used by the National Geographic society from 1922 to 1988.
Cons: slanting Alaska and Siberia, Greenland is still bigger than Australia.

 

[BrobDingnag]

Robinson until I die, or until some advanced machine learning algorithm manages to come up with a prettier one. Collignon is good for a laugh, though.

 

[Crazy Boris]

Kavrayskiy VII gang

Generally low distortion, it's easy to navigate, and it just generally looks nice

 

[BobTheBarbarian]

Robinson until I die, or until some advanced machine learning algorithm manages to come up with a prettier one. Collignon is good for a laugh, though.

Kavrayskiy VII gang
Generally low distortion, it's easy to navigate, and it just generally looks nice

I like these too. Of the three more common maps of this type (Robinson, Winkel Tripel, and Kavrayskiy VII), I prefer K-7 because it feels like a compromise between the other two. Robinson doesn't leave much space for Greenland and the Canadian arctic islands while it looks like Winkel Tripel distorts Australia and western North America more. Winkel Tripel also has curved latitude lines like Van der Grinten, but I think the latter does a better job with shape.

 

[Miranda Brawner]

Any of the family of equal-area cylindrical maps. Accompanied by a perspective projection that shows roughly what you would see from a point in space.

 

[Miranda Brawner]

Collignon is good for a laugh, though.

I just looked it up and I love it. If only we could somehow combine it with Goode-Homolosine and the World in a Tetrahedron.

 

[Hugh Third of Five]

Any of the family of equal-area cylindrical maps. Accompanied by a perspective projection that shows roughly what you would see from a point in space.

Equal-area pure cylindrical maps tend to have massive distortions of length and width. Things are stretched vertically near the equator and compressed vertically/stretched horizontally towards the poles unless you use the really weird orientations like oblique.

My favorite is the Mollweide. It's an equal area pseudocylindrical projection. Although there is greater distortion as you reach the longitudinal edges of the map.

It depends on what you're looking for. There will always be some distortion (unless you use a globe). The equal-area maps have zero area distortion, but that means they're going to have distortion in other aspects, like direction and shape. At the other end, the Mercator has zero distortion with respect to direction, but a lot of distortion with respect to area.

 

[Aghasverov]

Calling all geography nerds!

Which map projection do you like the best? Either for its geometric properties or for aesthetic appeal? Why is it your favorite?

Which would you use as a reference map to teach geography to others?

I like Van der Grinten I:

Pros: decent at maintaining shape of countries, less distortion than Mercator, was the official projection used by the National Geographic society from 1922 to 1988.
Cons: slanting Alaska and Siberia, Greenland is still bigger than Australia.

I like the VdG II, just for the added finesse that the parallels and meridians meet at right angles

 

[Aghasverov]

Equal-area pure cylindrical maps tend to have massive distortions of length and width. Things are stretched vertically near the equator and compressed vertically/stretched horizontally towards the poles unless you use the really weird orientations like oblique.

My favorite is the Mollweide. It's an equal area pseudocylindrical projection. Although there is greater distortion as you reach the longitudinal edges of the map.

It depends on what you're looking for. There will always be some distortion (unless you use a globe). The equal-area maps have zero area distortion, but that means they're going to have distortion in other aspects, like direction and shape. At the other end, the Mercator has zero distortion with respect to direction, but a lot of distortion with respect to area.

Always had a fondness for the Mollweide... also for the Hammer/Aitoff. There's just something about world maps in an ellipse with a 1:2 aspect ratio that "looks right"

 

[Aghasverov]

I like these too. Of the three more common maps of this type (Robinson, Winkel Tripel, and Kavrayskiy VII), I prefer K-7 because it feels like a compromise between the other two. Robinson doesn't leave much space for Greenland and the Canadian arctic islands while it looks like Winkel Tripel distorts Australia and western North America more. Winkel Tripel also has curved latitude lines like Van der Grinten, but I think the latter does a better job with shape.

Yeah, they're all good, but one thing annoys me a little about the "flat-polar", semi-elliptical pseudocylindrical projections - there's so damned many of them
There's practically an infinite number of ways to construct 'em, they're all compromises of some sort or another, and most of them are fairly attractive...

 

[Aghasverov]

This is an "old classic", maybe getting a little dated, but still fairly amusing:

 

[Gustone]

aitoff, sinusoidal

 

[Aghasverov]

aitoff, sinusoidal

I like the sinusoidal because of its remarkable mathematic simplicity - start with the equator and a central meridian half as long (2:1 ratio), and determine the length of your other parallels simply by multiplying by the cosine of each degree of latitude. Thus at 60N and 60S, your parallels are exactly half the length of the equator - as they should be. Voila! a map... As an additional bonus, it works out to be equal-area too...

 

[erictom333]

This is an "old classic", maybe getting a little dated, but still fairly amusing:

I was going to post that!
Back on topic: Plate Carrée, because:
a) equirectangular, so meridians are straight lines: useful for drawing alternate US states
b) actually shows the poles, unlike Mercator which alwats crops Antarctica off (I have a map of the world hangaing on my bedroom wall, but it's a Mercator map, and cuts off most of Antarctica.)
c) makes mathematical sense; maps θ and φ to x and y, anything else is too complicated
d) not too obscure

 

[Aghasverov]

I was going to post that!
Back on topic: Plate Carrée, because:
a) equirectangular, so meridians are straight lines: useful for drawing alternate US states
b) actually shows the poles, unlike Mercator which alwats crops Antarctica off (I have a map of the world hangaing on my bedroom wall, but it's a Mercator map, and cuts off most of Antarctica.)
c) makes mathematical sense; maps θ and φ to x and y, anything else is too complicated
d) not too obscure

The Plate Carree is underrated

 

[Aghasverov]

I think the irritating thing about Goode's Homolosine isn't the "orange-peel" technique - I can see the potential applications of that - it's the fact that it's not "a" projection but two projections grafted together - Sinusoidal in the equatorial band and Mollweide further N and S... giving that distinct "break" at about 40-44 N and S. That's just... cheating
The only thing I like about "Gall-Peters" is that when Arno Peters originally (re-)published it, he shifted the central meridian east away from Greenwich, to the "Meridian of Florence", which prevents the eastern extremity of Siberia from getting chopped off and thrown in on the other side of the world... that always annoyed me

 

[BobTheBarbarian]

The Plate Carree is underrated

I believe Tom van Sant's "Geosphere" map uses this projection. I got confused between it and the Miller projection because they look so similar. I suppose you can tell the difference from the stretched latitude lines in the Miller.

 

[Aghasverov]

Not familiar with the "Geosphere" map, but yeah, it looks like a Plate Carree. If I remember correctly, Miller's essentially takes the basic principle of the Mercator and reduces the North-South "stretching" by multiplying by 0.8...

 

[Aghasverov]

Wanted to *bump* this, as there's an infinite number of possible projections out there, bound to be some more opinions on this one!

Here's a link to a nifty little (very little, around 5MB) freeware program that I remembered from my college days:

MicroCAM for Windows

MicroCAM for Windows.

www.csiss.org

It's ancient, as am I but I checked the d/l link yesterday and it still works. Some newer versions of Windoze might be like "wtf is this?", but as I'm a semi-Luddite who refuses to use anything past 7, I wouldn't know

You can do fun stuff like try to show an entire hemisphere in the gnomonic projection (gaaaaah!)
Enjoy!

 

[BobTheBarbarian]

I was thinking (and maybe this has been done before), but I was wondering what would this look like:

What if you used a computer program to design a flat rectangular map based on the trigonometric positions of all other points on Earth relative to an arbitrary center?

Say for example, you started at zero latitude and zero longitude in the Gulf of Guinea, measured the great circle distance on the globe between that point and any other location on Earth, and then plotted that distance as a straight line on a flat map. Obviously, the only restriction would be that in creating this map you wouldn't be able to plot distances over the poles (say, from 0,0 to the other side of Antarctica over the South Pole). From there the process can continue until all landmasses have been accounted for.

How accurate would this be?