Antarctica looks like the opposite of a desert. Almost every photograph shows snow, glaciers or ice stretching beyond the horizon. Beneath that white surface lies enough frozen water to reshape coastlines around the world.
Yet desert is a measure of what falls from the sky, not how much water is already stored on the ground. By that definition, Antarctica is not merely a desert. At roughly 14 million square kilometres, it is the largest on Earth.
The distinction becomes clearest on the high Antarctic Plateau. Annual snow accumulation there is commonly equivalent to less than 50 millimetres of liquid water. That is a smaller water input than many hot deserts receive.
The kilometres of ice below are not evidence of heavy snowfall today. They are the result of cold, time and retention. A small surplus can become an enormous store when summer repeatedly fails to remove it.
A desert is defined by scarcity
The everyday image of a desert is sand, bare rock and punishing heat. Climate classifications are less visual. They identify deserts by persistent shortage of precipitation, often considered alongside how readily water could evaporate.
Nothing in that definition requires warmth. Polar deserts are places where water is scarce in a form that living things can use, even when frozen water covers the landscape. Antarctica is cold enough to hold ice and simultaneously too dry to receive much new moisture.
The continent is also far larger than the Sahara. The Antarctic Ice Sheet alone extends across almost 14 million square kilometres, according to the National Snow and Ice Data Center. Calling Antarctica the largest desert therefore refers to the scale of its dry climate, not to an ice-free patch within it.
The famous McMurdo Dry Valleys make the idea visible because they contain broad areas of exposed ground. Strong winds, low humidity and minimal snowfall prevent much permanent ice from forming there. Most of the Antarctic desert is harder to recognise because its accumulated ice conceals the rock.
The interior is much drier than the coast
“Antarctica receives less than 50 millimetres a year” is useful shorthand only if the location is stated. The figure applies to the elevated interior plateau, not uniformly to every part of the continent.
The Australian Antarctic Program’s climate summary estimates average accumulation across Antarctica at about 150 millimetres of water equivalent each year. On the plateau it falls below 50 millimetres. Near the coast it generally exceeds 200 millimetres, and one area near the Bellingshausen Sea receives more than 1,000 millimetres.
That gradient matters. Moist maritime air encounters the continent around its margins, where storms and rising terrain can produce substantial snow. Air travelling inland loses moisture along the way. By the time it reaches the high centre, it is extremely cold and dry.
Most Antarctic precipitation falls as snow or tiny ice crystals, although rain can occur near the coast. Even “snowfall” is not always easy to measure. Fierce winds lift old snow from the surface and mix it with new crystals, making a gauge incapable of reliably telling which material just fell from a cloud.
Fifty millimetres does not mean five centimetres of snow
The plateau figure is normally given as water equivalent. Researchers are expressing how deep the precipitation would be if it were melted into liquid, not reporting the depth of the loose snow layer.
Fresh snow contains abundant air, so 50 millimetres of water can arrive as a much greater depth of snow. The conversion is not fixed. Crystal shape, wind packing, temperature and previous surface conditions all affect snow density.
There is another complication: precipitation and accumulation are not identical. Precipitation is material delivered from the atmosphere. Accumulation is the amount left after snow has been redistributed by wind or lost through melting, runoff and sublimation, the direct change from ice to water vapour.
Some plateau surfaces gain almost nothing in a given year. Wind may scour one place and build a drift somewhere else. Researchers therefore combine stakes, snow pits, radar, ice cores, weather models and satellite observations to estimate the larger pattern.
Why the coldest continent is so dry
Cold air can sustain far less water vapour than warm air. Antarctica’s interior is both extremely cold and exceptionally high, with the South Pole standing about 2,830 metres above sea level on the ice sheet.
Moisture-bearing systems are most active around the coast. As air rises over the outer slopes, it cools and sheds snow. The remaining air reaching the interior has already lost much of its water. Persistent subsiding air over the plateau also discourages cloud formation.
This combination gives the interior clear skies, intense radiative cooling and very limited snowfall. The high surface makes the atmosphere colder still, while the bright snow reflects much of the Sun’s energy.
Gravity then sends dense surface air downhill. These katabatic winds can accelerate towards the coast, moving snow even when no storm is occurring. The Antarctic surface is not a quiet ledger on which every flake remains where it landed.
How a trickle becomes kilometres of ice
An ice sheet begins wherever snowfall survives the following summer. A new layer buries the old one. Pressure rearranges the crystals, reduces the air spaces and turns old snow into firn, an intermediate material denser than snow but still more porous than glacial ice.
Further burial compresses the firn into ice. Some air becomes sealed into bubbles, preserving samples of the ancient atmosphere. The National Snow and Ice Data Center’s account of ice-sheet formation describes this as the repeated survival and compression of snowfall over thousands of years.
A simple thought experiment shows how patience changes the scale. An average net gain equivalent to 50 millimetres of water each year would amount to 50 metres of water in a millennium if none were lost. Continue the process for tens of thousands of years and the total becomes substantial, although a real ice sheet also compacts, thins and flows.
The title’s reference to thousands of years captures the accumulation mechanism, but Antarctica’s full glacial history is far longer. Permanent glaciation began around 34 million years ago, and a 2024 reconstruction reported by the British Antarctic Survey indicates that the first major ice was concentrated in East Antarctica before expanding towards West Antarctica millions of years later.
The ice is thick, but it is not stationary
The Antarctic Ice Sheet has a mean thickness of about 2.16 kilometres and reaches a known maximum of 4,776 metres, according to the Australian Antarctic Program. Only a small fraction of the continent’s rock surface is ice-free.
It would be misleading, however, to picture every year’s snow remaining stacked directly above the place it fell. Ice deforms under its own weight. It spreads outward from the high interior and drains through glaciers and fast-moving ice streams towards the ocean.
Near the coast, snowfall is only one part of the mass balance. Ice can calve into the sea or melt from below where relatively warm ocean water reaches floating shelves. A previous ScienceBlog report on glacial earthquakes and changing flow at Thwaites Glacier showed how active the continent’s seaward margins can be.
The ice sheet survives when additions across its surface broadly balance the ice removed at its edges and base. More snow in one region does not automatically mean the entire sheet is growing, just as a dry year at one station does not describe the whole continent.
Snow can disappear without melting
The high interior sees little surface melting, but cold does not eliminate every route by which snow can be lost. Sublimation allows frozen water to enter the atmosphere directly. Wind can carry crystals into another basin or eventually out to sea.
These processes help produce hard wind glaze, sculpted dunes and blue-ice areas where old ice reaches the surface. In some locations, net accumulation is close to zero or negative despite temperatures that rarely approach melting.
That movement has unexpected consequences. ScienceBlog previously explored how flow and surface loss concentrate meteorites in Antarctic blue-ice fields. The same landscape that preserves ancient ice can also expose material carried within it.
“The snow rarely melts” is therefore accurate for the high plateau, but it should not be read as “the snow never leaves.” An ice sheet is a slow circulation system, gaining frozen water from above and losing it through several pathways.
A frozen reservoir built in a dry climate
Antarctica’s ice contains roughly 30 million cubic kilometres of frozen water. Its scale can make the desert label feel like wordplay, but the two facts describe different parts of the water budget.
The ice is a reservoir accumulated under conditions that strongly limit melting. The desert classification describes the small annual supply arriving from the atmosphere, especially across the plateau. A bank account can hold a vast balance while receiving only modest deposits if withdrawals remain small for long enough.
That balance is now changing unevenly. A warmer atmosphere can carry more moisture and may increase snowfall in some places, while warmer ocean water and faster ice flow increase losses elsewhere. Present-day ice-sheet change cannot be inferred from precipitation alone.
The apparent paradox is real only if desert is mistaken for heat or absence of ice. Antarctica is a continent where water is abundant as an inheritance and scarce as annual income. Its high interior receives only a thin yearly contribution, but cold preserves part of that contribution, pressure turns it to ice and gravity slowly carries it back towards the sea.
