Mercator Projection Explained: How It Compares With Peters, Robinson and Web Mercator
Source: Wikimedia Commons
Map Projections & Cartography

Mercator Projection Explained: How It Compares With Peters, Robinson and Web Mercator

Mercator’s 1569 map made rhumb lines straight for sailors and blew up the size of everything near the poles. Here is how it stacks up against equal-area, compromise and web projections.

Geography Worlds
March 19, 2026
Updated October 13, 2026
7 min read

The Mercator projection is a cylindrical, conformal world map that the Flemish cartographer Gerardus Mercator published in 1569 for navigators. Its one great property is that a line of constant compass bearing (a rhumb line) is a straight line on the map, so a sailor can draw a course with a ruler and steer it with a compass. The price is size: the map stretches everything away from the equator, by a factor of 4 in area at 60° latitude and about 33 at 80°, so Greenland looks as large as Africa even though Africa is about 14 times bigger, and the poles cannot be shown at all. That trade-off, perfect local shapes and angles against badly wrong areas, is what separates Mercator from every alternative below.

What Mercator drew in Duisburg in 1569, and why sailors needed it

Mercator (1512–1594) was born Gerard de Kremer in Rupelmonde, Flanders, and lived in Duisburg, in the Duchy of Cleves, from 1552 until his death. In 1569 he announced his projection on a world map 202 by 124 cm, printed on 18 sheets, under the title Nova et Aucta Orbis Terrae Descriptio ad Usum Navigantium Emendate Accommodata, “a new and more complete representation of the terrestrial globe properly adapted for use in navigation”. The map carries more than 5,000 words of text in fifteen legends, and the title alone shows he knew exactly what he had made.

The problem he was solving had been set out by the Portuguese mathematician Pedro Nunes, who described the rhumb line, or loxodrome, and in 1537 proposed a nautical atlas built from equirectangular sheets. Mercator had already engraved rhumb lines on a terrestrial globe in 1541. He never explained how he constructed the 1569 map; the first accurate tables came from the English mathematician Edward Wright in 1599, expanded in his 1610 Certaine Errors in Navigation, and Henry Bond published a mathematical formula around 1645, although Thomas Harriot had worked out the mathematics, unpublished, from about 1589. Mercator was not entirely alone in the idea either: the German polymath Erhard Etzlaub engraved small “compass maps” of Europe in 1511 that the projection scholar John Snyder judged in 1987 to use the same projection, before revising that in 1993 to “a similar projection”. Mercator’s other lasting legacy came later: his Atlas of 1595, published the year after his death, was the first time the word “atlas” appeared in connection with maps, though Mercator meant it as the name of a treatise on the cosmos rather than a book of maps.

Hand-coloured engraved portrait of a bearded Gerardus Mercator in a black cap, holding dividers over a globe
Gerardus Mercator, engraved by Frans Hogenberg in 1574 when the cartographer was 62, shown measuring a globe with dividers. Photo: Frans Hogenberg, Public domain, via Wikimedia Commons

Sailors were slow to adopt it. Without a way to fix longitude at sea, and with compasses pointing to magnetic rather than true north, a perfect chart was of limited use. Only in the middle of the 18th century, once the marine chronometer existed and magnetic declination had been mapped, did the Mercator become the standard navigation chart. Practically every printed marine chart still uses it today.

Mercator against the globe: Greenland, Africa and the 1/cos rule

Imagine wrapping a cylinder around the globe so that it touches along the equator. Every cylindrical projection stretches the parallels east–west so they are all as long as the equator; at latitude φ that stretch is 1/cos φ. Mercator’s trick was to stretch north–south by exactly the same amount at every latitude, which keeps small shapes and every angle correct. The cost is that areas grow by the square of that factor. At 60° north, roughly the latitude of Oslo, Helsinki and Anchorage, distances are doubled and areas quadrupled. Above 70° the map becomes practically unusable, and at the poles the stretch is infinite.

Mercator is often confused with the central cylindrical projection, made by shining an imaginary light from the Earth’s centre onto the cylinder. That version also blows up towards the poles, but it does not stretch north–south in step with east–west, so it is not conformal and its rhumb lines are not straight. Mapmakers can also shrink a Mercator map slightly so that scale is true along two chosen parallels instead of the equator. On such a secant version the scale is slightly too small between the two parallels and still grows without limit beyond them, so the basic trade-off does not change.

Equator (0°)1×
30°1.3×
45°2×
60°4×
70°8.5×
80°33×
How much the Mercator projection inflates area at each latitude, calculated as 1/cos² of the latitude. Source: Mercator projection formulas, as given in Wikipedia’s “Mercator projection”

The examples are familiar. Greenland appears the size of Africa, yet Africa is 14 times larger; Greenland’s real area is closer to that of the Democratic Republic of the Congo. Alaska looks as big as Australia, which is 4.5 times larger, and as big as Brazil, which is nearly 5 times larger. Madagascar and Great Britain look similar, though Madagascar is more than twice the size. Antarctica, stretched along the whole bottom edge, looks endless. The French mathematician Nicolas Auguste Tissot, in work published in 1859 and 1871, devised a way to show this: draw identical small circles on the globe and project them. On a Mercator map the circles stay perfectly round, proving the map is conformal, but they swell towards the poles.

World map on the Mercator projection with red circles of equal real size that grow larger towards the top and bottom of the map
Tissot’s indicatrices on a Mercator world map: every red circle covers the same area on the globe, but they swell with latitude while staying round. Photo: Stefan Kühn, CC BY-SA 3.0, via Wikimedia Commons

Mercator himself did not intend the map as a picture of relative size; for that he used the equal-area sinusoidal projection. But by the late 19th and early 20th centuries the Mercator had become perhaps the most common world map in schools and offices, and critics such as George Kellaway and the economist Irving Fisher argued it was unsuitable for general use. Atlases largely dropped it for world maps in the 1940s.

Mercator vs Gall–Peters and Equal Earth: shape against area

The sharpest contrast is with equal-area projections, which keep every country’s size correct and give up shape instead. The Scottish clergyman James Gall described one in 1855 and published it in 1885. The German historian Arno Peters independently devised the same construction in 1967 and presented it in 1973 as the “Peters world map”, arguing that conventional cartography reflected a Eurocentric view of the world. The American cartographer Arthur Robinson coined the name Gall–Peters in a 1986 pamphlet. Its standard parallels sit at 45° north and south, so Africa and South America look correctly sized but stretched tall and thin, while Canada and Russia look squashed.

Satellite-imagery world map on the Gall–Peters projection with Africa and South America elongated vertically
The world on the Gall–Peters equal-area projection: sizes are true, but tropical continents are drawn tall and narrow. Photo: Strebe, CC BY-SA 3.0, via Wikimedia Commons

Cartographers were unimpressed by both sides. In 1989 and 1990 seven North American geographical organisations adopted a resolution against using any rectangular world map, Mercator or Gall–Peters, for general purposes. When Boston Public Schools adopted Gall–Peters in March 2017, three cartographers, Bojan Šavrič, Bernhard Jenny and Tom Patterson, responded by designing Equal Earth (2018), an equal-area projection with curved sides that keeps continents better proportioned. On it, Greenland correctly appears at about 7% of Africa’s land area. In 2025 the African Union backed a campaign, led by Togo’s foreign minister Robert Dussey, for Equal Earth to replace Mercator in public use. Our explainer on the Peters projection follows that controversy in more detail.

Mercator vs Robinson and Winkel tripel: the National Geographic compromises

Between the two extremes sit compromise projections, which are neither conformal nor equal-area but keep every kind of distortion moderate. Arthur Robinson designed his projection in 1963 at the request of Rand McNally, working, by his own account, backwards from the shapes that looked right to the formulas that produced them. The National Geographic Society adopted the Robinson projection for its world maps in 1988, replacing the Van der Grinten, and in 1998 switched to the Winkel tripel, devised by the German cartographer Oswald Winkel in 1921 as the average of the equirectangular and Aitoff projections. “Tripel” refers to its aim of limiting three distortions at once: area, direction and distance.

Against these, the Mercator wins only on angles and straight bearings, and loses on almost everything a reader of a wall map cares about: the relative size of countries, the true shape of high-latitude coasts, and any view of the poles. That is why compromise projections dominate atlases, and why the Winkel tripel is often recommended for general world maps.

Mercator vs Web Mercator: Google Maps since 2005

The Mercator made an unexpected comeback online. When Google Maps launched in 2005 it adopted a variant now called Web Mercator (EPSG:3857), and Bing Maps, OpenStreetMap, Mapbox, Esri and almost every other online map followed. The reason is local: because Mercator is conformal, north is up everywhere and angles are locally correct, so street corners keep their true shape at every zoom level. Services such as Google Maps cut coverage off at about 85.05° north and south, which lets the tiling system show most of the world at the lowest zoom as a single square image.

Web Mercator is not quite the classic projection. It applies the formulas for a sphere to coordinates measured on the WGS 84 ellipsoid, which makes it very slightly non-conformal. Confusing it with true Mercator during coordinate conversion can shift positions by as much as 40 km, and the US National Geospatial-Intelligence Agency has declared it unacceptable for official use. Navigators still use true Mercator, and Mercator projections were also important in the mathematical development of plate tectonics in the 1960s. Its transverse form, turned on its side, underlies the Universal Transverse Mercator grid used for accurate regional mapping.

Whether the Mercator actually warps people’s mental picture of the world is less settled than often claimed. Perception studies over the past two decades have repeatedly failed to find a strong “Mercator effect”. The distortion itself is not in dispute, though, and for the reasons online maps still use it, see our guide to the projection behind Google Maps.

Recently updated guides