What happens when we actually measure the distortion?
Look at almost any familiar world map and one visual impression immediately jumps out: countries near the top and bottom of the map look enormous.
Greenland looks astonishingly large. Russia dominates northern Eurasia. Canada stretches across an enormous portion of the globe. Meanwhile, Africa—despite being one of Earth's largest landmasses—doesn't look nearly as overwhelming.
But how much of this is real geography, and how much is the mathematics of the map?
To answer that, I took the actual geographic boundaries of the countries in the Natural Earth 10m Admin-0 Countries dataset and calculated their areas after projecting them into Web Mercator. I then compared those areas with their corresponding spherical areas. Natural Earth describes its 10m dataset as a detailed global country dataset and currently lists 258 country/map units.
The results are remarkable.
First: what exactly is Mercator doing?
The Mercator projection was designed for navigation. Its great strength is that it preserves angles and local shapes, making a constant compass bearing appear as a straight line on the map.
That property is extremely useful for navigation.
But there is a price.
Mercator does not preserve area.
The local linear scale factor of the spherical Mercator projection is
where is latitude.
Because area has two dimensions, the local area scale factor is
This means that the further you move from the equator, the more dramatically the map enlarges an area.
At the equator:
so there is essentially no area inflation.
At 30° latitude:
so a small region appears about 33% larger.
At 60°:
so the same-sized region appears four times as large.
At 70°:
The distortion therefore becomes enormous toward the poles.
And this is not merely a theoretical curiosity. It profoundly changes how large countries appear.
The experiment: measure the actual countries
Rather than taking the latitude of a country's capital, its centroid, or some arbitrary representative latitude, we can do something much better.
We can take the entire geographic polygon of each country, transform it into Web Mercator, and calculate its projected area.
For comparison, we calculate its area on a sphere.
The resulting quantity is:
where is the Mercator inflation factor.
An inflation factor of:
1.0 = no area inflation
1.2 = 20% larger
2.0 = twice as large
5.0 = five times as large
10.0 = ten times as large
This is much more informative than simply saying that "Mercator exaggerates areas."
The results are extraordinary
Using the Natural Earth 10m polygons, the largest inflation factors are:
| Country/region | Mercator inflation | Apparent increase |
|---|---|---|
| Greenland | 16.42× | +1,542% |
| Norway | 9.02× | +802% |
| Iceland | 5.60× | +460% |
| Finland | 5.46× | +446% |
| Canada | 5.15× | +415% |
| Russia | 4.90× | +390% |
| Sweden | 4.87× | +387% |
| Estonia | 3.70× | +270% |
| United Kingdom | 2.93× | +193% |
| United States | 2.29× | +129% |
| France | 1.97× | +97% |
| China | 1.59× | +59% |
| Japan | 1.60× | +60% |
| India | 1.19× | +19% |
| Australia | 1.25× | +25% |
| Brazil | 1.06× | +6% |
These values are calculated from the actual polygons, so they account for the fact that a country can span a very large range of latitudes.
The striking result is Greenland.
Its polygon has a Mercator inflation factor of approximately 16.4.
In other words, if you simply measured its area on a Web Mercator map, you would obtain an area more than sixteen times its spherical surface area.
That is why Greenland can look comparable in size to Africa on familiar world maps even though Africa is vastly larger. The well-known comparison is that Africa is roughly 14 times the area of Greenland in reality.
India provides a useful baseline
India is particularly interesting because most of the country lies in the tropical and subtropical latitudes.
The calculation gives:
So India is inflated by approximately 19.1% on Web Mercator.
That is substantial—but modest compared with countries occupying high northern latitudes.
Consider Russia:
Russia's area is therefore inflated by approximately 390%.
The ratio between the two inflation factors is:
So, relative to India, the Mercator projection gives Russia roughly 4.1 times as much projection-induced area exaggeration.
That is an important distinction.
It does not mean Russia is four times India's area.
It means that Mercator's distortion acts approximately four times more strongly on Russia than on India.
Now make India = 1
This is where the comparison becomes intuitive.
India's actual area is approximately one India.
On an equal-area map, the relative areas are approximately:
| Region | Equal-area size relative to India |
|---|---|
| India | 1.00 |
| China | ~2.98 |
| USA | ~3.00 |
| Russia | ~5.39 |
| Europe | ~7.3 |
| Africa | ~9.5 |
So the real geographic ranking is approximately:
Africa > Europe > Russia > USA ≈ China > India
But the map projection changes our visual intuition dramatically.
Mercator's distortion disproportionately enlarges high-latitude regions.
That is why northern countries can occupy far more visual real estate than their actual area warrants.
Africa is the fascinating counterexample
Africa is perhaps the most useful demonstration of why looking only at the map can be misleading.
Africa is enormous.
Its area is roughly:
Yet much of Africa lies close to the equator.
Consequently, the Mercator area inflation is relatively modest compared with Russia, Canada or Greenland.
This produces a fascinating reversal in visual perception.
A high-latitude landmass receives a large Mercator boost.
An equatorial landmass receives comparatively little.
So Mercator does not simply make the entire world "bigger."
It changes the relative visual importance of different latitudes.
The countries least affected are near the equator
The opposite end of the calculation is equally revealing.
Among the least inflated Natural Earth polygons are:
| Country | Inflation factor |
|---|---|
| Nauru | 1.0001× |
| São Tomé and Príncipe | 1.0001× |
| Singapore | 1.0006× |
| Gabon | 1.0007× |
| Equatorial Guinea | 1.0010× |
| Uganda | 1.0011× |
| Ecuador | 1.0012× |
| Rwanda | 1.0013× |
| Kenya | 1.0016× |
This makes intuitive sense.
The closer an area is to the equator, the closer the Mercator scale factor is to 1.
Countries around the equator therefore get something very close to their true area.
Why Greenland is such an extreme example
Greenland is the perfect demonstration of what happens when a huge landmass occupies high latitudes.
Mercator keeps stretching the north-south direction as latitude increases.
Consequently, a large portion of Greenland receives an enormous area multiplier.
The result is an inflation factor of roughly:
That doesn't mean the map is "wrong" in the sense of being mathematically defective.
It means that the projection was designed to preserve a different property—angles and navigation, rather than area.
This distinction is crucial.
Mercator isn't a bad map
It is tempting to conclude that Mercator is simply a bad or misleading projection.
That would be unfair.
Mercator solved an important problem.
For navigation, preserving angles is extremely valuable. A line of constant compass bearing—the famous rhumb line—can be represented as a straight line on a Mercator chart.
That is why Mercator remained enormously important in nautical navigation.
The problem arises when we use a projection optimized for navigation to answer questions about area.
It would be like using a thermometer to measure distance.
The instrument isn't "bad."
We're asking it the wrong question.
Enter Equal Earth
This is where Equal Earth becomes useful.
Equal Earth is an equal-area pseudocylindrical projection designed specifically for world maps. Unlike the Robinson projection, it preserves relative areas.
That means:
on an Equal Earth map is the same as the corresponding true-area ratio.
The shapes are necessarily changed somewhat—the fundamental problem of representing a spherical Earth on a flat sheet cannot be escaped—but the relative areas are preserved.
So if your question is:
"How large is Africa compared with India?"
an equal-area projection is the appropriate tool.
If your question is:
"How do I plot a constant-bearing navigation route?"
Mercator may be the better tool.
What does this mean in the real world?
The implications go beyond geography textbooks.
1. Our mental map of the world is partly constructed by projection
For generations, many people have learned geography from rectangular world maps.
The visual size of a country becomes part of our mental model of the world.
That means projection isn't merely a technical detail.
It can influence our intuition about:
geographic scale
distance
population distribution
environmental extent
geopolitical importance
resource distribution
The distortion is especially consequential when comparing regions at very different latitudes.
2. Africa's size is routinely underestimated visually
Africa is one of the clearest examples.
On a Mercator-style map, Africa does not look nearly as dominant as its actual area warrants.
An equal-area projection immediately changes that perception.
This is particularly important in education because children often learn geography through visual comparison before they encounter numerical area measurements.
3. Northern countries acquire a visual advantage
Russia, Canada, Greenland and much of northern Europe occupy very high latitudes.
Their land receives substantial Mercator enlargement.
That doesn't mean Mercator "makes them more important."
But it does mean that visual comparisons of their physical size become unreliable.
A map reader can easily come away with the impression that northern countries occupy more of Earth's surface than they actually do.
4. It matters for environmental maps
Consider maps showing:
forests
agricultural land
deserts
protected areas
carbon storage
biodiversity
climate zones
land-use change
If the underlying visualization uses a non-equal-area projection, visual comparisons of coloured regions can be misleading.
For quantitative spatial analysis, the appropriate projection—or an equal-area calculation performed before visualization—is therefore important.
5. It matters for communicating climate and environmental change
Suppose two regions contain the same number of square kilometres of forest.
If one is near the equator and another is at high latitude, their apparent areas can be dramatically different on Mercator.
A viewer might therefore interpret the visual size of the coloured region as an indication of the magnitude of the environmental phenomenon.
That can be wrong.
For area-based communication, equal-area projections are often much more defensible.
But there is an even deeper lesson
The most interesting thing about this exercise is not that Mercator distorts maps.
We already knew that.
The interesting thing is how strongly the distortion varies among countries.
The calculation gives a continuum:
The map is therefore effectively assigning different "visual currencies" to different latitudes.
One square kilometre near the equator occupies approximately one square kilometre's worth of Mercator map area.
One square kilometre near 60°N occupies roughly four times as much.
At 70°N, it occupies more than eight times as much.
That is the mathematical reason our visual intuition can become so badly disconnected from geographic reality.
And there is a final twist
We often say:
"Africa looks too small on Mercator."
That's true, but incomplete.
The more precise statement is:
Africa looks relatively small because high-latitude landmasses are enlarged much more strongly than equatorial landmasses.
This is a subtle but important distinction.
Africa isn't necessarily being "shrunk."
Rather, the rest of the world is being enlarged by different amounts.
That is why comparing countries visually on a Mercator map can produce such surprising results.
The R code
The entire analysis can be reproduced in R using the Natural Earth shapefile.
The code below assumes that you have downloaded and unzipped the Natural Earth 10m Admin-0 Countries dataset.
############################################################
# MERCATOR AREA DISTORTION FOR EVERY COUNTRY
#
# This script:
# 1. Reads Natural Earth country polygons
# 2. Calculates their spherical surface areas
# 3. Projects them into Web Mercator
# 4. Calculates projected areas
# 5. Calculates the Mercator inflation factor
# 6. Ranks countries from most to least distorted
#
# Required packages:
# sf
# s2
# dplyr
# ggplot2
############################################################
# ----------------------------------------------------------
# 1. Load packages
# ----------------------------------------------------------
library(sf)
library(s2)
library(dplyr)
library(ggplot2)
# ----------------------------------------------------------
# 2. Read the Natural Earth 10m country polygons
# ----------------------------------------------------------
# Change this path to wherever you extracted the
# Natural Earth shapefile.
countries <- st_read(
"ne_10m_admin_0_countries.shp",
quiet = TRUE
)
# ----------------------------------------------------------
# 3. Remove Antarctica
# ----------------------------------------------------------
# Web Mercator cannot represent the poles.
# Its Y coordinate tends toward infinity as latitude
# approaches ±90 degrees.
#
# Antarctica is therefore excluded from this particular
# Mercator calculation.
countries <- countries %>%
filter(CONTINENT != "Antarctica")
# ----------------------------------------------------------
# 4. Calculate spherical geographic area
# ----------------------------------------------------------
# Web Mercator is mathematically defined on a sphere.
#
# We therefore want a spherical reference area rather than
# the ellipsoidal WGS84 area normally returned by sf.
#
# s2 calculates areas on a sphere.
#
# The radius cancels when we calculate the ratio between
# projected and reference areas, so the important quantity
# here is the relative area.
spherical_area <- s2_area(
st_as_s2_geography(st_geometry(countries))
)
# Convert square metres to square kilometres.
spherical_area_km2 <- spherical_area / 1e6
# ----------------------------------------------------------
# 5. Project the polygons into Web Mercator
# ----------------------------------------------------------
# EPSG:3857 is Web Mercator, the projection widely used
# by web mapping systems.
countries_mercator <- st_transform(
countries,
crs = 3857
)
# ----------------------------------------------------------
# 6. Calculate projected Mercator area
# ----------------------------------------------------------
# st_area() now calculates ordinary planar polygon area
# because the polygons are in a projected coordinate system.
mercator_area <- st_area(countries_mercator)
mercator_area_km2 <- as.numeric(mercator_area) / 1e6
# ----------------------------------------------------------
# 7. Calculate the Mercator inflation factor
# ----------------------------------------------------------
# This is the central calculation.
#
# A value of:
#
# 1.0 = no inflation
# 1.5 = 50% larger
# 2.0 = twice the true area
# 5.0 = five times the true area
#
# Because the entire polygon is used, this automatically
# incorporates the fact that countries can span many
# different latitudes.
countries$true_area_km2 <- spherical_area_km2
countries$mercator_area_km2 <- mercator_area_km2
countries$mercator_inflation <-
countries$mercator_area_km2 /
countries$true_area_km2
# ----------------------------------------------------------
# 8. Calculate percentage inflation
# ----------------------------------------------------------
countries$inflation_percent <-
(countries$mercator_inflation - 1) * 100
# ----------------------------------------------------------
# 9. Sort from most to least distorted
# ----------------------------------------------------------
results <- countries %>%
st_drop_geometry() %>%
select(
NAME,
ADMIN,
CONTINENT,
true_area_km2,
mercator_area_km2,
mercator_inflation,
inflation_percent
) %>%
arrange(desc(mercator_inflation))
# Display the 20 most distorted countries.
head(results, 20)
# ----------------------------------------------------------
# 10. Find India
# ----------------------------------------------------------
india <- results %>%
filter(NAME == "India")
india
# ----------------------------------------------------------
# 11. Compare every country with India's distortion
# ----------------------------------------------------------
india_factor <- india$mercator_inflation
results <- results %>%
mutate(
relative_to_india =
mercator_inflation / india_factor
)
# ----------------------------------------------------------
# 12. Save the complete results
# ----------------------------------------------------------
write.csv(
results,
"mercator_inflation_all_countries.csv",
row.names = FALSE
)
# ----------------------------------------------------------
# 13. Plot the most distorted countries
# ----------------------------------------------------------
top20 <- results %>%
slice_head(n = 20) %>%
mutate(
NAME = reorder(NAME, mercator_inflation)
)
ggplot(
top20,
aes(
x = NAME,
y = mercator_inflation
)
) +
geom_col() +
coord_flip() +
labs(
title = "Mercator inflation factor",
subtitle = "Natural Earth 10m country polygons",
x = NULL,
y = "Mercator area / spherical area"
) +
theme_minimal()
# ----------------------------------------------------------
# 14. Plot selected countries
# ----------------------------------------------------------
selected <- results %>%
filter(
NAME %in% c(
"India",
"China",
"United States of America",
"Russia",
"Canada",
"Greenland",
"Brazil",
"Australia",
"South Africa",
"Norway"
)
) %>%
mutate(
NAME = reorder(NAME, mercator_inflation)
)
ggplot(
selected,
aes(
x = NAME,
y = mercator_inflation
)
) +
geom_col() +
coord_flip() +
labs(
title = "How strongly does Mercator enlarge different countries?",
x = NULL,
y = "Mercator inflation factor"
) +
theme_minimal()
One technical detail matters
There are several different ways of calculating "true area," and they should not be mixed casually.
Web Mercator uses a spherical mathematical model, whereas geographic datasets such as WGS84 describe Earth using an ellipsoid.
For the purpose of this analysis, the cleanest approach is to compare the Web Mercator projected area with the corresponding spherical reference area.
The absolute area of the sphere is not the important quantity here.
The important quantity is the ratio:
because this isolates the projection's area distortion.
The same calculation could alternatively be performed using an ellipsoidal geodesic area, but then a small additional difference would arise from comparing an ellipsoidal Earth with a spherical projection model.
The bigger lesson
Maps don't merely show the world.
They mathematically transform the world.
Every flat map makes compromises. Some preserve angles. Some preserve area. Some preserve distances along particular lines. Some attempt to balance several kinds of distortion.
There is no perfect flat map of a spherical planet.
The important question is therefore not:
"Which map is correct?"
It is:
"Correct for what?"
For navigation, Mercator remains extraordinarily useful.
For comparing the physical size of continents and countries, an equal-area projection such as Equal Earth is much more appropriate. Equal Earth was specifically designed as a visually appealing world-map projection that nevertheless preserves relative area.
And once you have actually calculated the distortion, the familiar world map starts to look very different.
Greenland isn't enormous.
Russia isn't as gigantic as it looks.
Canada isn't nearly as large as its Mercator silhouette suggests.
And Africa really is enormous.
The mathematics was hiding in plain sight.
Data and reproducibility
The calculations in this article use the Natural Earth 10m Admin-0 Countries dataset, version 5.1.1. Natural Earth describes this as its detailed country-level dataset and notes that the default boundaries represent de facto territorial control.
The projection used for the distortion calculation is Web Mercator (EPSG:3857).
The equal-area comparison discussed above uses the Equal Earth projection, which is explicitly classified as an equal-area projection.
