Sea level rise: 33 years of satellite measurement

Since October 1992, a single orbit has been flown without interruption by a lineage of altimetry satellites: Topex/Poseidon, then Jason-1, Jason-2, Jason-3 and today Sentinel-6 Michael Freilich. Every ten days they measure the distance separating them from the surface of the ocean, everywhere between 66 degrees of latitude north and south. Thirty-three years of this measurement have produced one of the most solid series in climatology, and it says two things: the sea has risen by about ten centimetres, and it is rising faster and faster.

11.6 cmof rise since January 1993

3.41 mm/anon average over the whole period

3.86 mm/anover the last ten years

1236measurements, one every ten days

Niveau moyen de l’océan mondial, 1993 à 2026

zone claire : incertitude publiée par le producteur, 2.8 mm sur la dernière mesure. Série corrigée du rebond post-glaciaire, signaux saisonniers retirés, lissage six mois. Hover over the curve to read a date.

A rise that is accelerating

Mean rate of sea level rise, by complete decade, global ocean.
Decade	Speed	Difference from the 1993-2003 decade
1993 à 2003	2.41 mm/an	reference
2003 à 2013	2.82 mm/an	+0.41 mm/an, soit +17 %
2013 à 2023	4.00 mm/an	+1.59 mm/an, soit +66 %

L’accélération mesurée sur l’ensemble de la période est de 0.075 mm/an². Autrement dit, chaque année qui passe, la mer monte un peu plus vite que l’année précédente.

Not all basins are rising at the same rate

Basin series not corrected for post-glacial rebound, and therefore comparable with each other but not with the global curve above, which is corrected. Click a legend entry to isolate a basin.

Rate of rise by basin, ranked by the last ten years. The rise is calculated from the trend, and not from the difference between the first and the last measurement, which the variability of the final column would make misleading.
Basin	Since 1993	Last ten years	Rise	Variability
Mediterranean	3.07 mm/an	5.94 mm/an	10.3 cm	± 19.0 mm
Indian Ocean	3.26 mm/an	4.72 mm/an	11.0 cm	± 5.3 mm
Océan mondial (corrected for rebound)	3.41 mm/an	3.86 mm/an	11.5 cm	± 2.5 mm
North Atlantic	3.55 mm/an	3.64 mm/an	11.9 cm	± 5.3 mm
South Pacific	2.97 mm/an	3.48 mm/an	10.0 cm	± 7.9 mm
North Pacific	3.31 mm/an	3.18 mm/an	11.1 cm	± 5.3 mm
South Atlantic	3.11 mm/an	2.48 mm/an	10.4 cm	± 5.8 mm

La dernière colonne donne l’écart-type autour de la tendance : c’est la marge dans laquelle le niveau fluctue d’une année sur l’autre, sous l’effet du vent, de la pression atmosphérique et des grandes oscillations climatiques. Elle vaut 19.0 mm en Méditerranée contre 2.5 mm pour l’océan mondial : un bassin semi-fermé bouge beaucoup plus vite que la moyenne planétaire, ce qui explique l’allure tourmentée de la courbe rouge.

Données : AVISO+ / ODATIS, indicateur Mean Sea Level, DOI 10.24400/527896/AVISO-2025.010, Licence Ouverte / Open Licence. Crédits : EU Copernicus Marine Service / CNES / LEGOS / CLS, aviso.altimetry.fr/msl. Dernière mesure du 16 August 2026, fichier mis à jour le 3 November 2026.

Mean sea level, in millimetres, relative to January 1993

Measurements by satellite altimetry, value taken at the start of each year. Global ocean corrected for post-glacial rebound, basins not corrected.

Year	Océan mondial	Mediterranean	North Atlantic	South Atlantic	North Pacific	South Pacific	Indian Ocean
1993	0.8 mm	-87.6 mm	-27.6 mm	11.5 mm	-24.2 mm	2.2 mm	3.3 mm
1994	8.8 mm	-7.2 mm	-7.5 mm	9.3 mm	6.2 mm	9.2 mm	-1.0 mm
1995	11.8 mm	3.4 mm	-5.1 mm	11.5 mm	4.7 mm	21.2 mm	5.5 mm
1996	12.3 mm	59.6 mm	1.9 mm	19.5 mm	4.6 mm	12.9 mm	12.2 mm
1997	14.7 mm	57.5 mm	11.2 mm	13.2 mm	9.9 mm	20.3 mm	5.5 mm
1998	18.6 mm	46.9 mm	11.2 mm	16.4 mm	9.3 mm	37.5 mm	11.5 mm
1999	15.6 mm	4.9 mm	16.3 mm	21.4 mm	16.2 mm	5.4 mm	9.8 mm
2000	22.4 mm	6.2 mm	18.5 mm	23.0 mm	19.5 mm	14.7 mm	22.7 mm
2001	26.2 mm	61.3 mm	20.1 mm	23.4 mm	26.4 mm	26.8 mm	16.7 mm
2002	29.1 mm	-5.4 mm	26.7 mm	30.1 mm	22.9 mm	28.1 mm	27.8 mm
2003	33.0 mm	40.3 mm	21.3 mm	24.7 mm	30.5 mm	43.1 mm	24.4 mm
2004	35.3 mm	31.5 mm	31.3 mm	38.7 mm	31.9 mm	37.4 mm	24.6 mm
2005	38.1 mm	1.8 mm	30.4 mm	45.1 mm	33.9 mm	40.5 mm	31.0 mm
2006	42.2 mm	12.7 mm	26.8 mm	43.4 mm	44.8 mm	43.9 mm	31.0 mm
2007	41.9 mm	15.5 mm	31.2 mm	37.8 mm	37.3 mm	42.9 mm	39.7 mm
2008	44.1 mm	12.7 mm	30.8 mm	34.9 mm	47.1 mm	34.7 mm	42.3 mm
2009	48.3 mm	40.7 mm	33.8 mm	45.1 mm	40.9 mm	40.3 mm	50.8 mm
2010	53.0 mm	115.0 mm	41.3 mm	49.5 mm	45.0 mm	56.0 mm	45.6 mm
2011	50.7 mm	84.1 mm	37.0 mm	48.0 mm	41.5 mm	39.4 mm	49.8 mm
2012	58.1 mm	9.5 mm	48.4 mm	47.5 mm	49.8 mm	49.6 mm	60.8 mm
2013	66.8 mm	69.9 mm	52.5 mm	56.5 mm	63.7 mm	61.0 mm	63.2 mm
2014	65.8 mm	46.9 mm	54.9 mm	63.4 mm	53.3 mm	58.2 mm	67.9 mm
2015	72.0 mm	40.8 mm	58.5 mm	60.6 mm	58.6 mm	74.3 mm	67.4 mm
2016	80.9 mm	66.7 mm	65.6 mm	67.6 mm	69.9 mm	84.8 mm	77.8 mm
2017	81.2 mm	31.9 mm	67.3 mm	71.0 mm	84.2 mm	78.3 mm	61.5 mm
2018	84.5 mm	66.8 mm	85.3 mm	88.8 mm	84.9 mm	72.0 mm	71.5 mm
2019	88.3 mm	72.5 mm	79.8 mm	87.1 mm	84.1 mm	90.3 mm	68.2 mm
2020	92.3 mm	75.8 mm	88.7 mm	98.3 mm	88.3 mm	81.7 mm	81.1 mm
2021	96.4 mm	106.7 mm	90.7 mm	90.3 mm	95.3 mm	79.9 mm	92.7 mm
2022	100.7 mm	64.3 mm	100.5 mm	100.8 mm	90.6 mm	90.0 mm	88.2 mm
2023	104.4 mm	100.9 mm	99.8 mm	97.8 mm	93.9 mm	96.6 mm	90.6 mm
2024	111.8 mm	134.0 mm	109.7 mm	102.4 mm	102.6 mm	102.8 mm	108.5 mm
2025	112.5 mm	97.2 mm	100.7 mm	91.0 mm	111.8 mm	100.5 mm	106.1 mm
2026	113.9 mm	136.6 mm	107.8 mm	99.6 mm	108.5 mm	98.6 mm	105.8 mm

What an altimetry satellite actually measures

The satellite sends a radar pulse downwards and times its return: from that it derives its height above the water. Its own position is known to within a centimetre thanks to a network of ground beacons. The difference between the two gives the height of the sea surface relative to the centre of the Earth, once the effects of the atmosphere, the waves and the tides have been removed.

This precision rests on a point of engineering: all these missions have followed exactly the same ground track since 1992, and each new satellite flies for a few months in formation with its predecessor, a few seconds apart, long enough to compare their measurements over the same waves. Without that overlap, every change of satellite would introduce a step of a few millimetres, indistinguishable from the signal being sought.

One useful point: this curve does not describe the shoreline. It gives the level of the ocean, not the depth of water in front of a given port, which also depends on how the local ground is rising or subsiding.

A rate that has almost doubled

This is the clearest result in the series. Between 1993 and 2003, the sea was rising by 2.41 mm per year. Between 2013 and 2023, by 4.00 mm per year, two thirds faster. Over the last ten rolling years, the rate is 3.85 mm per year. The acceleration measured over the whole period reaches 0.075 mm per year per year, which amounts to saying that each passing decade adds about 0.75 mm per year to the rate of rise.

These figures are recalculated here from the producer’s raw series, and they land on the values it publishes on its own side: 3.67 mm per year since 1999 in our calculation, against 3.6 plus or minus 0.3 mm per year announced; 0.75 mm per year per decade of acceleration, against 0.8 plus or minus 0.4. The reconstruction is therefore faithful.

The Mediterranean, the fastest of the last decade

Over the period as a whole, the Mediterranean is rising slightly more slowly than the global mean. But over the last ten years it is the fastest rising basin of all: 5.88 mm per year, against 3.85 for the global ocean. For a sea that holds the Pelagos Sanctuary and a resident population of fin whales and sperm whales, the figure deserves to be followed.

It also deserves to be handled with caution. A semi-enclosed basin reacts far faster than an ocean to winds and atmospheric pressure: the Mediterranean level fluctuates from one year to the next within a margin of nearly 20 mm around its trend, against 2.5 mm for the global mean. That is what gives its curve its tormented look, and it is why a single decade is not enough to conclude that there is a lasting runaway. Our page on the cetaceans of the Mediterranean sets out what is seen there.

Focus: the Bay of Biscay

The bay is not one of the data producer’s reference basins: it has to be pieced together from the grid of trends, grid cell by grid cell. The result is worth the detour. Over 493 grid cells at sea, the sea there is rising by 4.38 mm per year, against 3.36 mm per year for the global mean of the same product over the same period. That is 30% faster than the global average.

4.38 mm/anin the bay, measured by satellite

+30 %faster than the global average

4.19 mm/anmesuré au sol par 7 marégraphes

218 ansof continuous records at Brest

Where the sea is rising fastest in the bay

Tendance par maille de 0,25 degré, 1999 à 2025, sur 493 mailles en mer. Les cases vides sont des terres, ce qui dessine le trait de côte. La hausse est remarquablement uniforme : de 3.71 à 4.91 mm/an seulement, pour un écart-type de 0.17 mm/an entre mailles. C’est le message à retenir de cette carte : à l’échelle du golfe, la mer monte partout à la même vitesse. Les nuances entre classes voisines sont du même ordre que l’incertitude d’une maille, elles ne dessinent pas de structure interprétable.

Brest, 218 ans de mesures

Moyennes annuelles, écart au niveau moyen de la période 1960-1990, en millimètres. Points bleus : chaque année. Courbe rouge : moyenne glissante sur onze ans. 2,310 mois de relevés, 88 % de la période couverte.

Rate of rise measured at Brest, by period. The sea there was level throughout the nineteenth century.
Period	Speed	Months measured
1807 à 1900	0.02 mm/an	945
1900 à 1950	2.25 mm/an	500
1950 à 2000	1.52 mm/an	561
1996 à 2026	4.07 mm/an	352

L’accélération mesurée à Brest depuis 1850 est de 1.41 mm/an par siècle.

Eight tide gauges, one and the same answer

Trend over the common period 1996-2025, with its two standard deviation interval. Coverage indicates the share of months actually measured.
Station	Since	Trend 1996-2025	Coverage
Arcachon-Eyrac	1967	4.89 ± 1.43 mm/an	74 %
La Rochelle-La Pallice	1941	4.11 ± 1.10 mm/an	92 %
Brest	1807	4.07 ± 1.06 mm/an	98 %
Le Conquet	1971	3.88 ± 1.03 mm/an	97 %
Concarneau	1971	3.82 ± 1.34 mm/an	87 %
Les Sables-d’Olonne	1965	3.76 ± 1.14 mm/an	94 %
Saint-Jean-de-Luz (Socoa)	1942	3.46 ± 1.32 mm/an	61 %
Saint-Nazaire	1941	3.43 ± 1.21 mm/an	96 %

Two instruments, two different measurements, one and the same result. Le satellite mesure le niveau de la mer par rapport au centre de la Terre ; le marégraphe le mesure par rapport au quai auquel il est fixé, lequel peut lui-même monter ou s’affaisser. Sur la fenêtre exacte du satellite, les marégraphes donnent 4.19 mm/an et l’altimétrie 4.38 mm/an : 0.19 mm/an d’écart, bien à l’intérieur des incertitudes de chaque station. Cet accord suggère un sol peu mobile dans le golfe, mais ne le démontre pas : il faudrait pour cela les vitesses verticales mesurées par GNSS, que nous n’exploitons pas ici.

Grille de tendances : AVISO+ / ODATIS, grille de tendances, DOI 10.24400/527896/AVISO-2025.010, période 1999-02-21 à 2025-08-02, non corrigée du rebond post-glaciaire. Marégraphes : Permanent Service for Mean Sea Level (PSMSL), moyennes mensuelles RLR, psmsl.org. Emprise retenue : 43.5 à 48.5 N, 8 à 1 O. Mise à jour du 3 November 2026.

Rate of sea level rise in the Bay of Biscay

Altimétrie satellitaire, moyenne sur le golfe : 4.38 mm par an entre 1999-02-21 et 2025-08-02, contre 3.36 mm par an pour la moyenne mondiale du même produit.

Year	Océan mondial	Mediterranean	North Atlantic	South Atlantic	North Pacific	South Pacific	Indian Ocean
1993	0.8 mm	-87.6 mm	-27.6 mm	11.5 mm	-24.2 mm	2.2 mm	3.3 mm
1994	8.8 mm	-7.2 mm	-7.5 mm	9.3 mm	6.2 mm	9.2 mm	-1.0 mm
1995	11.8 mm	3.4 mm	-5.1 mm	11.5 mm	4.7 mm	21.2 mm	5.5 mm
1996	12.3 mm	59.6 mm	1.9 mm	19.5 mm	4.6 mm	12.9 mm	12.2 mm
1997	14.7 mm	57.5 mm	11.2 mm	13.2 mm	9.9 mm	20.3 mm	5.5 mm
1998	18.6 mm	46.9 mm	11.2 mm	16.4 mm	9.3 mm	37.5 mm	11.5 mm
1999	15.6 mm	4.9 mm	16.3 mm	21.4 mm	16.2 mm	5.4 mm	9.8 mm
2000	22.4 mm	6.2 mm	18.5 mm	23.0 mm	19.5 mm	14.7 mm	22.7 mm
2001	26.2 mm	61.3 mm	20.1 mm	23.4 mm	26.4 mm	26.8 mm	16.7 mm
2002	29.1 mm	-5.4 mm	26.7 mm	30.1 mm	22.9 mm	28.1 mm	27.8 mm
2003	33.0 mm	40.3 mm	21.3 mm	24.7 mm	30.5 mm	43.1 mm	24.4 mm
2004	35.3 mm	31.5 mm	31.3 mm	38.7 mm	31.9 mm	37.4 mm	24.6 mm
2005	38.1 mm	1.8 mm	30.4 mm	45.1 mm	33.9 mm	40.5 mm	31.0 mm
2006	42.2 mm	12.7 mm	26.8 mm	43.4 mm	44.8 mm	43.9 mm	31.0 mm
2007	41.9 mm	15.5 mm	31.2 mm	37.8 mm	37.3 mm	42.9 mm	39.7 mm
2008	44.1 mm	12.7 mm	30.8 mm	34.9 mm	47.1 mm	34.7 mm	42.3 mm
2009	48.3 mm	40.7 mm	33.8 mm	45.1 mm	40.9 mm	40.3 mm	50.8 mm
2010	53.0 mm	115.0 mm	41.3 mm	49.5 mm	45.0 mm	56.0 mm	45.6 mm
2011	50.7 mm	84.1 mm	37.0 mm	48.0 mm	41.5 mm	39.4 mm	49.8 mm
2012	58.1 mm	9.5 mm	48.4 mm	47.5 mm	49.8 mm	49.6 mm	60.8 mm
2013	66.8 mm	69.9 mm	52.5 mm	56.5 mm	63.7 mm	61.0 mm	63.2 mm
2014	65.8 mm	46.9 mm	54.9 mm	63.4 mm	53.3 mm	58.2 mm	67.9 mm
2015	72.0 mm	40.8 mm	58.5 mm	60.6 mm	58.6 mm	74.3 mm	67.4 mm
2016	80.9 mm	66.7 mm	65.6 mm	67.6 mm	69.9 mm	84.8 mm	77.8 mm
2017	81.2 mm	31.9 mm	67.3 mm	71.0 mm	84.2 mm	78.3 mm	61.5 mm
2018	84.5 mm	66.8 mm	85.3 mm	88.8 mm	84.9 mm	72.0 mm	71.5 mm
2019	88.3 mm	72.5 mm	79.8 mm	87.1 mm	84.1 mm	90.3 mm	68.2 mm
2020	92.3 mm	75.8 mm	88.7 mm	98.3 mm	88.3 mm	81.7 mm	81.1 mm
2021	96.4 mm	106.7 mm	90.7 mm	90.3 mm	95.3 mm	79.9 mm	92.7 mm
2022	100.7 mm	64.3 mm	100.5 mm	100.8 mm	90.6 mm	90.0 mm	88.2 mm
2023	104.4 mm	100.9 mm	99.8 mm	97.8 mm	93.9 mm	96.6 mm	90.6 mm
2024	111.8 mm	134.0 mm	109.7 mm	102.4 mm	102.6 mm	102.8 mm	108.5 mm
2025	112.5 mm	97.2 mm	100.7 mm	91.0 mm	111.8 mm	100.5 mm	106.1 mm
2026	113.9 mm	136.6 mm	107.8 mm	99.6 mm	108.5 mm	98.6 mm	105.8 mm

Period	Brest, mm per year
1807 à 1900	0.02
1900 à 1950	2.25
1950 à 2000	1.52
1996 à 2026	4.07

What the Brest tide gauge says

Brest holds one of the longest sea level series in the world: monthly readings since 1807, that is 218 years and 2,310 months of measurement, covering 88% of the period. That historical depth gives a perspective thirty-three years of satellite data cannot provide.

And it says this: throughout the nineteenth century, the sea did not rise at Brest. The trend over 1807-1900 is 0.02 mm per year, in other words strictly nil. The movement begins at the turn of the twentieth century, reaches 1.3 mm per year on average over that century, and 4.07 mm per year over the last thirty years. This is not an old trend continuing, it is a phenomenon that started, then accelerated.

Two independent instruments, one same figure

A satellite in orbit and a float attached to a quay have nothing in common, neither the measurement principle, nor the sources of error, nor even the quantity measured: the first locates the surface relative to the centre of the Earth, the second relative to the ground, which also moves. Comparing them is therefore a real test.

Over the exact window of the satellite grid, the tide gauges of the bay give 4.19 mm per year on average, altimetry 4.38. Nineteen hundredths of a millimetre per year of difference, well below the uncertainty of each station. The eight tide gauges retained also agree with one another, from 3.4 to 4.9 mm per year depending on the port, around a mean of 3.9 over 1996-2025.

What this changes on that coast

Here again, caution is called for: four millimetres per year do not trouble a rorqual out beyond the shelf. The effect is coastal, and it is twofold.

First, a storm now arrives on a higher starting level. Flooding does not depend on the mean level but on the surge that adds to it. And at Brest, the mean of the 2015-2025 decade exceeds that of the 1895-1905 decade by 24 centimetres. That much of a head start given to every gale. Xynthia, in 2010, struck the Pertuis Charentais on that already raised base.

Second, transitional environments are shifting. The bay has two large estuaries, the Gironde and the Loire, and one semi-enclosed basin, Arcachon, where common bottlenose dolphins regularly come in. A higher sea pushes salt further upstream, redraws the mudflats and the seagrass beds, and squeezes intertidal habitats against the sea walls that border them. Those are where the fish nurseries lie, and so, at the end of the chain, the resource of the predators.

Our other pages on this area give the rest of the picture: plankton and temperature, cetacean sightings, fishing and collisions.

Where this water comes from

For the 1993-2018 period, the IPCC attributes 46% of the rise to the thermal expansion of sea water, 44% to the melting of land ice (mountain glaciers, Greenland, Antarctica) and 10% to transfers of water stored on the continents.

The first figure is the most counter-intuitive: close to half the rise in the waters is not added water, it is the same water taking up more room because it is warmer. The ocean in fact absorbs more than 90% of the excess heat accumulated by the climate system. Sea level is therefore, for half of it, a thermometer.

Eleven centimetres, and why they count all the same

Let us say it plainly: eleven centimetres of extra water mean nothing to a whale on the open sea. A pelagic species does not perceive the mean level of the ocean, and it would be dishonest to present this curve as a direct threat to cetaceans.

Its interest lies elsewhere, and it is twofold.

It is first of all an indicator of the heat stored by the ocean. Half of this curve measures the expansion of the water, and therefore its warming. And it is that warming, not the level, which displaces the plankton, then the fish, then the predators that follow them. Our monitoring of plankton and temperature in the Bay of Biscay and of ocean currents shows this mechanism at a more legible scale.

It is then a real pressure on coastal and estuarine species. River and estuary dolphins live precisely where the rise in the waters translates into salt moving upstream and into the transformation of deltas and mangroves. The Irrawaddy dolphin, the tucuxi and freshwater cetaceans in general occupy narrow habitats, which the sea nibbles away from below while dams and water abstraction reduce them from above.

Three limits to keep in mind

The measurement stops at 66 degrees of latitude. That is the constraint of the reference orbit. In other words, this series says nothing about the Arctic or the approaches to Antarctica, that is to say precisely the waters of the beluga, the narwhal and the bowhead whale. For those regions, other missions and other series are needed.

Two different corrections coexist. The global curve is corrected for post-glacial rebound, that slow deformation of the ocean basins inherited from the last glaciation, which weighs about 0.3 mm per year. The basin series are not, since they are not published that way. That is why the two graphs above are kept separate: mixing them would give the global figure an artificial advantage over the basins.

A short trend is not worth a long one. Seasonal signals have been removed from the series, but year-to-year variability remains, carried by the large climate oscillations such as El Niño. Over three or four years, that variability largely outweighs the underlying signal: this is why only complete decades appear in the acceleration table.

Sources

Data: the Mean Sea Level indicator from AVISO+ and the ODATIS pole, referenced on Ifremer’s Sextant catalogue, DOI 10.24400/527896/AVISO-2025.010, under the Licence Ouverte. Credits required by the producer: EU Copernicus Marine Service / CNES / LEGOS / CLS, aviso.altimetry.fr/msl. Breakdown of the contributions to the rise: IPCC, Sixth Assessment Report, Working Group I, chapter 9. The trends, accelerations and standard deviations shown on this page are recalculated by us from the public series, and checked against the values published by the producer.