Every Map Is Wrong

1Gnomonic

Around 580 BC (traditionally) · Azimuthal · Straight lines are shortest routes

Tradition credits Thales of Miletus with using this projection for star maps, though nothing he wrote about it survives. Picture a light at the very center of a glass globe, shining the coastlines onto a flat sheet touching the surface. For centuries it was mostly a tool for designing sundials, and it was called the "horologium." The lines of a sundial for a given latitude match the meridians on a gnomonic map centered there.

Its one superpower is rare and useful: every straight line on a gnomonic map is a great circle, the shortest path between two points on the globe. The line from New York to London on the map below is the route a plane would actually fly. Navigators drew their route on a gnomonic chart, then transferred it to a Mercator chart to work out compass headings. The price is that it can't show even half the world, and distortion explodes toward the edges.

2Stereographic

2nd century BC · Azimuthal · Conformal

The stereographic projection is often credited to the Greek astronomer Hipparchus, though the oldest surviving text about it is Ptolemy's Planisphaerium, written about 300 years later. Its first big job was the astrolabe, the handheld star computer used by astronomers and sailors for centuries. It works for that because it maps every circle on the sky to a circle on the flat plate.

It's also conformal: shapes are correct in any small area. In fact, it's the only perspective projection (one you could make with a light and a screen) that keeps shapes right. Today it maps the poles, where the military's Universal Polar Stereographic grid takes over from UTM north of 84°N and south of 80°S.

3Orthographic

2nd century BC · Azimuthal · The view from space

The orthographic projection is how the Earth looks from very far away: one hemisphere, curving away at the edges like a globe. Hipparchus is said to have used it for astronomical calculations, and the Egyptians may have known it before him. Early writers called it the "analemma." The name "orthographic" came in 1613.

It doesn't preserve area, shape or distance. It just looks real, which is its whole point. During World War II, the illustrator Richard Edes Harrison drew striking orthographic "global" maps for the American public. Ours is centered on Chicago.

4Equirectangular (Plate Carrée)

Around AD 100, possibly earlier · Cylindrical · The simplest map there is

Draw latitude and longitude as a plain square grid and you have the equirectangular projection. Ptolemy credited Marinus of Tyre with it around AD 100, and Snyder thought it might go back to Eratosthenes three centuries earlier. Marinus's version wasn't quite square, though, and Ptolemy wasn't impressed. He called it "a manner of representing the distances which gives the worst results of all."

It survives because it's simple. Sixteenth-century sailors used it as their "plane chart" until Mercator's map caught on. Today it's the default for global satellite imagery and climate data, since every pixel lines up directly with a latitude and longitude. If you've ever plotted raw GPS coordinates without choosing a projection, this is the map you made.

5Azimuthal Equidistant

Ancient; oldest surviving map 1426 · Azimuthal · True distance and direction from the center

On this map, the distance and direction from the center point to anywhere else on Earth are both exactly right. It may go back to Egyptian star charts. The oldest one that survives is a 1426 star map by Conrad of Dyffenbach, and Mercator put it in the polar corners of his famous 1569 map.

You've seen it on the United Nations emblem, which shows the world as an azimuthal equidistant map centered on the North Pole. It was designed in 1945, and that's the version shown here. Aviation-era maps used it centered on a single city, to show true distances from there to everywhere else. The far side of the world gets stretched into a ring around the edge, which is why Antarctica looks like a circular wall.

6Mercator

1569 · Cylindrical · Conformal · Straight compass headings

Gerardus Mercator's 1569 world map has a long Latin title that ends, roughly, "corrected for use in navigation." That was the point. On his map, a line of constant compass heading (a rhumb line) is straight. A sailor could draw a line from port to port, read the angle, and steer that heading the whole way. To make that work, Mercator had to stretch the map north and south more and more toward the poles, which is why the poles never appear at all.

Mercator probably spaced his latitude lines by hand, since the math tables he needed didn't exist yet. The Englishman Edward Wright published the exact numbers in 1599. On the map below, the dashed line from New York to London is the constant heading, and the solid line is the actual shortest route. Notice how the circles stay round but balloon toward the poles.

Using it for general world maps was, in one historian's words, "a use unintended by Mercator." It's still the standard for nautical charts, which is what it was built for. The trouble came when publishers hung it on classroom walls. On Mercator, Greenland looks slightly bigger than Africa, which is 14 times its size.

7Sinusoidal

By 1570 · Pseudocylindrical · Equal-area

Nobody knows who invented the sinusoidal projection. The earliest map that uses it is a 1570 world map by Jean Cossin of Dieppe. It's been named after two later users, Nicolas Sanson and John Flamsteed, neither of whom created it, and it was even called "Mercator Equal-Area" after appearing in editions of Mercator's atlas.

The idea is beautifully simple. Every line of latitude is drawn at its true length, so the map gets narrower toward the poles and areas come out exactly right. The cost is shape: the far edges of the map are badly sheared. It's still at work today. NASA stores most of its high-resolution MODIS satellite land data on a sinusoidal grid, because equal-area tiles make global statistics simpler.

8Bonne

1752 (much older roots) · Pseudoconic · Equal-area · The heart

This one is named for Rigobert Bonne, a French cartographer who used it for a 1752 atlas of the French coast, but its ideas go back to Ptolemy. Its ancestor is the heart-shaped Werner projection from around 1514. Bonne's version centers the map on the area being drawn, which keeps that area accurate.

The lines of latitude are arcs of circles, all at their true length, and areas are exactly right. France's topographic maps switched to it in 1803, and atlases used it for maps of individual continents into the 20th century. Shown as a whole world, it becomes a heart.

9Lambert Conformal Conic

1772 · Conic · Conformal

In 1772 the mathematician Johann Heinrich Lambert published seven new projections in a single work, and three of them are in this list. This was the first. Lambert was curious whether the stereographic and the Mercator, which look nothing alike, were really two ends of one family of shape-preserving maps. They are, and this projection sits in between.

Then it sat almost unused for 140 years. The French used a version of it for battle maps in World War I, and the US Coast and Geodetic Survey revived it in 1918. Today it's everywhere. US aeronautical charts use it, so do many State Plane zones (the official coordinate systems for each US state), and it's ideal for regions that are wider east to west than north to south.

measured over the lower 48

10Lambert Azimuthal Equal-Area

1772 · Azimuthal · Equal-area

The last major projection in Lambert's 1772 book keeps areas exactly right while keeping directions from the center true. Of the classic azimuthal maps, it's the one whose scale changes least as you move away from the center.

It's the official map of European statistics. The European Union's standard grid for comparing data across countries, known as EPSG:3035, is a Lambert azimuthal equal-area projection centered at 52°N, 10°E, in northern Germany. If you've seen EU maps of population or land use, there's a good chance they were drawn on it. GIS people can spot its coordinates by their oddly familiar offsets: 4,321,000 and 3,210,000 meters.

measured over Europe

11Transverse Mercator (UTM)

1772 · Cylindrical · Conformal · The workhorse

Take Mercator, turn it on its side, and you get the transverse Mercator. Lambert described it in the same 1772 book. Mercator is most accurate along the equator. The transverse version is most accurate along one north-south line, so it works brilliantly for a narrow north-south strip.

The 1911 Encyclopaedia Britannica left it out as "seldom used." Then Gauss and later Louis Krüger worked out the precise math for the real, slightly squashed Earth, and in 1947 the US Army built the Universal Transverse Mercator (UTM) system on it: 60 zones, each 6° wide, covering the world from 84°N to 80°S. Snyder called it probably the most-used projection of all for precise mapping. The USGS's current US Topo maps use it. The dashed lines here mark UTM zone 16, the zone Chicago is in. Inside the strip, distortion is tiny. Outside it, the circles start to swell.

measured inside UTM zone 16

12Albers Equal-Area Conic

1805 · Conic · Equal-area

Heinrich Christian Albers, a German, published this projection in 1805. It's built for regions in the middle latitudes that are wide east to west, which describes the United States almost perfectly.

The US Coast and Geodetic Survey's Oscar Adams promoted it in 1927, calling it "as good as any other and in many respects superior to all others." The USGS adopted it for national maps, and today's standard version for the lower 48 (EPSG:5070) is what you're looking at. If you've seen a map of the US where the northern border curves gently, that's probably Albers. It's the right choice whenever you're comparing areas across states, like acres of farmland or forest cover.

measured over the lower 48

13Mollweide

1805 · Pseudocylindrical · Equal-area

Carl Mollweide, a mathematician in Halle, published this in 1805, the same year as Albers. It fits the whole world into an ellipse while keeping every area correct. It went mostly unnoticed until the French physicist Jacques Babinet revived it in 1857.

Astronomers love it. NASA's famous maps of the cosmic microwave background, the afterglow of the Big Bang, use Mollweide, because an equal-area map is the fair way to show a whole sky. It also inspired several later projections on this list, including Goode's.

14Polyconic

Around 1820 · Polyconic · America's survey map

Ferdinand Hassler, the first head of the US Survey of the Coast, began promoting this projection around 1820 for one huge job: mapping the American coastline. His idea was to treat each line of latitude as the edge of its own cone. Each map sheet could then be built from simple tables, and every sheet would be accurate near its own center.

It worked well enough that the US Geological Survey used it for its topographic maps from the 1880s until the late 1950s. Hassler himself died in 1843 after a fall while trying to protect his instruments in a storm. The projection's weakness shows up when you stretch one version too far. A 1928 USGS manual warned that it was "not at all suitable for a single-sheet map of the United States," even though people did it anyway, as the curved edges here show.

measured over North America

15Gall-Peters

1855 / 1973 · Cylindrical · Equal-area · The controversy

In 1855, a Scottish minister named James Gall presented an equal-area world map with tall, stretched-looking continents. He published it properly in 1885. It got little attention.

In 1973, the German historian Arno Peters called a press conference in Bonn to announce what he presented as a revolutionary new map, the fairest picture of the world, which showed the global south at its true size instead of Mercator's shrunken version. It was essentially Gall's projection. Cartographers objected. They pointed out that it wasn't new and that it distorts shapes about as badly as Mercator distorts sizes. Peters's map caught on anyway. In 1989 a group of North American geography organizations, National Geographic among them, urged publishers to stop using rectangular world maps altogether, Mercator and Peters. In 2017, Boston Public Schools switched to Gall-Peters maps, and the news coverage led directly to the last projection on this list.

16Van der Grinten

1904 · Compromise · The whole world in a circle

Alphons van der Grinten was a Chicagoan. He invented this projection in 1898 and patented it in 1904, aiming to blend the familiar look of Mercator with the rounded edges of Mollweide. The whole world fits inside a circle.

National Geographic made it the standard for its world maps in 1922, according to the Society itself (Snyder's manual says 1943), and kept it until 1988. That's a lot of classroom walls. The catch is the polar regions. When National Geographic finally dropped it, the Society said its Van der Grinten maps showed Greenland 554% too large.

17Winkel Tripel

1921 · Compromise · National Geographic's current choice

The German cartographer Oswald Winkel created this in 1921 by averaging two other projections. "Tripel" isn't a person. It's German for "triple," usually explained as an effort to keep three kinds of distortion (area, direction and distance) all low at once instead of eliminating just one.

National Geographic began using it in the mid-1990s and made it the standard for its world maps by 1998, replacing Robinson. A 2007 study by the astrophysicists David Goldberg and J. Richard Gott, which scored world maps on several kinds of error, found that Winkel Tripel "has low distortion on most measures and excellent quality overall." Its poles are drawn as lines, and the edges curve gently, which most people find natural-looking.

18Goode Homolosine

1923 · Pseudocylindrical, interrupted · Equal-area · The orange peel

J. Paul Goode, a geography professor at the University of Chicago, took the orange-peel problem literally. If you have to tear the peel to flatten it, why not choose where to tear? In 1923 he joined the sinusoidal projection (good near the equator) to the Mollweide (good at higher latitudes) where their scales match, and cut the result along the oceans so each continent gets its own center line.

The result is an equal-area map where the continents keep their shapes remarkably well, at the cost of the oceans being sliced apart. ("Homolosine" combines homolographic, an old name for Mollweide, with sine.) It also anchored Goode's World Atlas for Rand McNally, and it was the most-used equal-area world map in American textbooks from 1940 to 1960.

19Robinson

1963 · Compromise · Designed by eye

In 1963 the map company Rand McNally asked Arthur Robinson, a geography professor at the University of Wisconsin, for a world map that simply looked right. He did something unusual. Instead of deriving it from a mathematical formula, he defined it with a table of coordinates, chosen so that the shapes and sizes looked right. He called it "orthophanic," meaning "right-appearing."

In 1988 National Geographic's cartographers reviewed twenty candidates and unanimously chose Robinson to replace Van der Grinten, cutting Greenland's exaggeration from 554% to 60%. Robinson was no stranger to big maps. During World War II, he had run the map division of the Office of Strategic Services.

20Equal Earth

2018 · Pseudocylindrical · Equal-area · The modern answer

When Boston's schools adopted Gall-Peters in 2017, three cartographers, Bojan Šavrič, Tom Patterson and Bernhard Jenny, decided the world deserved a better equal-area map. Their goal was a map with every area correct that still looked as pleasant as Robinson. They published Equal Earth in 2018, built from simple equations so computers can draw it quickly.

It caught on fast. NASA's climate scientists at the Goddard Institute for Space Studies use it for their global temperature maps, and mapping software added it soon after it was published. Then came the 2026 UN resolution, which singles out Equal Earth as a projection that "more accurately represents relative size." So the oldest argument in cartography got a new answer, at least for the question of size.