Sea level has risen 23 centimetres since 1901

The world’s oceans stand about 23 centimetres higher than they did in 1901¹. That is the global average: the height of the whole sea surface, measured by tide gauges in harbours until the early 1990s and by satellites since.

It doesn’t sound like much. But the sea has risen faster since 1900 than in any century in at least 3,000 years², and the rise is bunched at the recent end. Nearly half of it, about 11 of the 23 centimetres, has come since 1993¹.

The older part of the record is the less certain part. The total since 1901 is known to within about 5 centimetres either way, while the rise since 1993 is pinned down to less than half a centimetre¹.

Why you may see 10 or 20 centimetres elsewhere

Look up how much the sea has risen and you’ll find several answers. They differ because each one counts from a different year.

NASA’s 10 centimetres starts in 1993, when satellites began measuring, and is rounded⁵; the average of the satellite records puts it nearer 11¹. The 20 centimetres from the Intergovernmental Panel on Climate Change (IPCC) covers 1901 to 2018², and the sea has risen almost 3 centimetres since¹.

The US National Oceanic and Atmospheric Administration (NOAA) gives 21 to 24 centimetres, counted from 1880 to 2023¹⁰. Our figure uses the IPCC’s starting year, 1901, and runs to the start of 2026. After that it moves along a projection, which “How this number was made” describes.

Sea level is rising about 4 millimetres a year, three times as fast as a century ago

Sea level is now rising by about 4 millimetres a year. The World Meteorological Organization (WMO) puts the last decade, 2016 to 2025, at 4.1 millimetres a year, nearly double the 2.1 of the first satellite decade⁸. From 1901 to 1971 the average was 1.3 millimetres a year².

A millimetre is thin, but this one is spread over every ocean on Earth. It takes about 362 billion tonnes of melted land ice to lift the sea by that much³.

Single years jump around. In 2024 the sea rose 5.9 millimetres, well above the 4.3 that NASA expected⁶. In 2025 it rose only 0.8⁵. So the rate you read depends on which years it is measured across, and recent published figures run from about 3.8 to 4.8 millimetres a year¹⁷.

The jumps come from the weather. In 2025 the Pacific was in its cool phase, La Niña, and unusually heavy rain over the Amazon basin moved water from the ocean onto land. That held the rise down even as the oceans warmed to a record. NASA expects the water to drain back within a year and the fast rise to return⁵.

People have held water back on land too. A sharp increase in the water stored behind new dams was the main reason the rise slowed in the 1970s¹¹.

Warm water expands, and melting ice adds more

Two things raise the sea. Water takes up more room as it warms (thermal expansion), and water arrives from the land as glaciers and ice sheets melt.

From 1971 to 2018, expansion caused about half the rise. Mountain glaciers added about 22%, Greenland 13% and Antarctica 7%. The last 8% or so was water once stored on land, in reservoirs and aquifers, that ended up in the sea³.

Both causes trace back to a warming planet, and people have been the main cause of the rise since at least 1970³.

The ice sheets’ share is growing. Since 2016 Greenland has supplied about 22% of the rise and Antarctica 14%³. The balance also shifts with the heat of a given year. In recent years about two-thirds of the rise came from melting ice and a third from expansion. In 2024 that flipped, and two-thirds came from the water itself warming and expanding⁶.

The sea keeps rising long after the warming stops

Sea level answers to temperature slowly. Heat takes centuries to work its way into the deep ocean, and an ice sheet takes longer still to melt. So the level we measure today is still catching up with warming that has already happened.

The IPCC expects another 10 to 25 centimetres by 2050 “whether or not greenhouse gas emissions are reduced”. Even if emissions reach net zero, the sea carries on rising, because the deep ocean keeps warming and the ice sheets keep adjusting³.

Further out the numbers get larger and less certain. Over the next 2,000 years the IPCC’s estimate, which it holds with low confidence, is a rise of 2 to 3 metres if warming is held to 1.5 °C, and 2 to 6 metres at 2 °C².

The planet has been this warm before. About 125,000 years ago it was 0.5 to 1.5 °C warmer than in the late 1800s, a range that today’s warming already sits inside, and the sea likely stood 5 to 10 metres higher than it does now².

The water would come from the ice. Greenland holds enough to raise the oceans 7.4 metres, West Antarctica 5.3 and East Antarctica 52.2, about 65 metres in all¹². That is the size of the store, not a forecast. The 23 centimetres so far is about a third of one percent of it.

On today’s policies the sea rises about 12 more centimetres by 2050, and half a metre by 2100

On today’s policies the sea rises roughly another 12 centimetres by 2050 and close to half a metre by 2100. That would put it about 36 centimetres above its 1901 level in 2050 and about 72 in 2100³.

These figures are the IPCC’s intermediate scenario (SSP2-4.5), which is also behind the CO2 projection because it tracks current policies most closely. The IPCC counts from the 1995 to 2014 average; here and below we have added the 15.6 centimetres the sea had risen by then, so every level is measured from 1901¹.

The range around the 2100 figure is wide: 60 to 92 centimetres on that one scenario. Across scenarios, the middle estimate for 2100 runs from about 54 centimetres on the lowest emissions path to about 93 on the highest³.

The rise also speeds up along the way, to 5.8 millimetres a year around mid-century and 7.7 by the century’s end on the intermediate scenario³. The satellites point the same way. If the acceleration they have recorded since 1993 simply continued, the sea would rise another 17 centimetres or so in the three decades after 2023, which is close to the IPCC’s middle path⁹.

Could the sea rise two metres this century?

A rise of two metres by 2100 is unlikely, but the IPCC says it cannot be ruled out. On the highest emissions scenario its likely range for 2100 is about 79 to 117 centimetres above the 1901 level. Beyond that, it says a rise “approaching 2 m by 2100 and 5 m by 2150” cannot be ruled out. It gives that statement low confidence, its label for a finding with little evidence behind it².

That tail rests on a process called marine ice cliff instability: once the floating ice in front of an ice sheet (an ice shelf) is gone, the tall cliffs left behind collapse under their own weight. Only one ice sheet model in the IPCC’s assessment included it³, and newer work has pushed back.

A 2022 expert assessment put the high end for 2100 at up to 1.6 metres. Its authors argued there is no evidence of an ice shelf break-up early and large enough to drive a major loss of Antarctic ice this century¹⁴. A 2024 study ran three models of Thwaites Glacier in West Antarctica with its floating ice removed, and in none of them did the cliffs keep retreating this century¹⁵.

West Antarctica is still the place to watch. Thwaites alone holds 65 centimetres of sea level. The international team studying it concluded in 2025 that a full collapse is unlikely in the next few decades, but that the glacier will retreat further and faster through this century and the next. Over that time, it said, the loss of the whole West Antarctic ice sheet, worth more than 3 metres, cannot be ruled out¹⁶.

The ocean around Thwaites is likely to warm at about three times its historical rate this century, whether emissions follow a middle path or the most ambitious climate targets, so more melting of the floating ice there looks unavoidable¹⁷.

Push the date out and the picture changes. By 2300, without any ice cliff collapse, the IPCC’s range is 1.7 to 6.8 metres on the highest emissions path and 0.3 to 3.1 on a low one, counted from the 1995 to 2014 average. That is why it says the main uncertainty about a high-end rise “is ‘when’ rather than ‘if’”³.

Higher seas already show in US high-tide floods and in Tuvalu

The sea doesn’t rise evenly. About two-thirds of the world’s coastline should see a rise within 20% of the global figure². Elsewhere the land itself is moving. Where it is still springing back from the weight of the last ice age, local sea level can fall. Sinking river deltas such as New Orleans and the Ganges–Brahmaputra see the water climb faster.

Ice has an odder effect. A melting ice sheet loses some of its gravitational pull on the water around it, so the sea drops nearby and rises most in the opposite hemisphere and the tropics³.

In the United States, NOAA’s latest outlook has coasts flooding at high tide, with no storm involved, on a median of 7 more days a year than in 2000. That is an increase of more than 250%, and NOAA projects 55 to 85 such days a year by 2050. Sinking land plays a part there too, as NOAA itself notes¹⁹.

In Tuvalu, in the Pacific, the sea is nearly 15 centimetres higher than 30 years ago. It is rising about 5 millimetres a year, one and a half times the global rate, while the islands sink at varying speeds²⁰.

About 680 million people, nearly one in ten as of 2010, live on coastal land less than 10 metres above the sea²¹. One study found that under high emissions, land that is home to up to 630 million people today would sit below the yearly flood line by 2100²². What that means for them is its own story: the effects of rising sea levels.

What cutting emissions still changes

Nothing in the record shows the rise slowing, and the pause in 2025 was weather.

The rise to mid-century is already set. After that the paths split, mainly in how fast the sea is rising. On the lowest emissions path the sea is still rising about 4 millimetres a year at the end of the century, much as it is now. On the highest it is rising 12³.

The gap in height follows later. By 2150 the middle estimates are about 84 centimetres above 1901 on a low emissions path and 148 on the highest³. One study found that every five years of delay before emissions peak adds about 20 centimetres to sea level in 2300¹⁸.

Coasts will have to adapt either way. What emissions seem to decide is the pace they have to keep up with: a sea rising about as fast as today’s, or one rising three times as fast.