Countries That Might Not Exist in Fifty Years (An inundation ledger — vanishing republics, drowning deltas, and the strange law being written so that a nation can outlive its own ground)

By September 2023, the Parliament of Tuvalu had amended the national constitution to declare that the country would exist in perpetuity, even if the physical territory of the islands went under. Tuvalu would remain a state, the text reads, “notwithstanding the impacts of climate change or other causes resulting in loss to the physical territory of Tuvalu.” Eleven thousand people were legislating for the possibility that they might one day have no ground to stand on.

Bookkeeping, not flourish. Over the twentieth century the global sea rose by around 7.5 inches, and the World Meteorological Organization now puts the floor at half a metre more by 2100, with the pessimistic tail running to two. Pacific waters have climbed at two to three times the global rate — almost thirty centimetres in the western basin since 1990 alone — and the reasons lie as much in wind and ocean circulation as in meltwater. Violent, erratic weather arrives over the top of that: storms out of season, king tides reaching the second row of houses, rain that comes all at once or not at all.

Fifty years is the window in which this stops being a story about adaptation and becomes a story about removal — not the disappearance of coastlines, which is old news and mostly survivable, but the disappearance of entire political entities. Tuvalu, Kiribati, the Marshall Islands and the Maldives are the obvious candidates, and they are not shy about saying so. Behind them sits a second tier whose exposure is larger in absolute terms and whose populations run into the tens of millions: the deltas of Bangladesh, Vietnam and Egypt, the low coast of Guyana, the reclaimed provinces of the Netherlands, the sinking north of Jakarta, the Nigerian lagoon, the Gulf coast of the United States. None will vanish. Several will lose the ground their economies stand on.

What follows moves from physics to particular countries, from salt in the well to a body of law that has, in the past two years, begun to harden in these nations’ favour, and finally to the places further inland that imagine they have time. Arithmetic this unpleasant is rarely this clear. Nobody who studies it seriously expects to be surprised. Whether the world looks away for another thirty years is another matter.

A Ledger at a Glance

Atoll states — where the highest natural ground is lower than a two-storey house

Country	Population	Land area	Highest natural point	Principal exposure	Practical horizon
Tuvalu	~11,000	26 km²	4.5 m (Funafuti)	Overwash, freshwater lens, king tides	Uninhabitable in parts by 2050–2070
Kiribati	~120,000	811 km²	3 m on Tarawa; 81 m on Banaba	Whole atoll chains, salinisation	Tarawa critically exposed by 2050
Marshall Islands	~42,000	181 km²	2–3 m on Majuro; 10 m on Likiep	Overwash, contaminated sites, storm surge	Majuro largely dependent on engineering by 2070
Maldives	~520,000	298 km²	2.4 m	Overwash, erosion, tourism infrastructure	Dependent on reclamation and defence
Tokelau (NZ)	~1,500	12 km²	~5 m	Total exposure, no higher ground	2050–2080
Nauru	~12,000	21 km²	71 m	Coastal ring only; interior safe	Coastal displacement only
Palau	~18,000	459 km²	242 m	Coastal villages, reefs	Partial
Federated States of Micronesia	~105,000	702 km²	791 m (Pohnpei)	Outlying atolls entirely; high islands partially	Atolls 2050–2080

Next tier — mainland and delta states where the exposure is counted in millions

Country	Population on exposed coast	Mechanism	Horizon
Bangladesh	~20 million in the coastal belt	Delta subsidence, cyclones, salinity	Displacement already under way
Vietnam	~17–21 million in the Mekong Delta	Subsidence from groundwater, sediment starvation	Much of the delta below sea level by 2100
Egypt	Nile Delta, 40% of national industry	Subsidence, sediment trapped behind the Aswan Dam	Alexandria and Port Said under chronic threat
Indonesia	North Jakarta and the Java north coast	Subsidence of up to 25 cm per year	Capital already relocated
Guyana	~90% of the population	Coastal plain half a metre to a metre below high tide	Permanent pumping and defence
Netherlands	~26% of land below sea level	Storm surge, river flooding	Managed by design, not by geography
Nigeria	Lagos and the Niger Delta	Subsidence, surge, erosion	Chronic, accelerating
Thailand	Bangkok and the upper Gulf	Subsidence, surge	Chronic
United States	Gulf and Atlantic seaboard	Subsidence, surge, nuisance flooding	Regional, accelerating

Sea Level Is Not a Tide

A tide goes out; the baseline does not. When a king tide floods a village in the Marshall Islands, the water recedes and the village dries, and a visitor could reasonably conclude that nothing much has changed. What has changed is the level from which the tide operates. Floods that used to arrive twice a decade now arrive twice a year, and events that were once-in-a-century will, on current trajectories, come round annually in the lifetime of children now in primary school.

Rise is not uniform, and the reasons are physical rather than political. Water expands as it warms, and the ocean has absorbed more than ninety per cent of the excess heat trapped by greenhouse gases; thermal expansion alone accounts for roughly a third of observed rise, melting glaciers for another fifth. Greenland and Antarctica, which between them hold enough ice to raise the sea some sixty-five metres if fully unlocked, supply the remainder and most of the uncertainty. Both are losing mass faster than they were a generation ago, and the Antarctic contribution is the largest single unknown in the calculation.

For 2100, the Intergovernmental Panel on Climate Change puts the likely global figure somewhere between 0.28 and 0.55 metres if emissions fall sharply, and between 0.63 and 1.01 metres if they do not. Two complications darken the picture. One is that these numbers describe a global mean, and a global mean is not what drowns a village: regional winds, the position of the great currents and the shape of the sea floor all shift local water levels by tens of centimetres in either direction, and in the western Pacific the stacking of warm water has pushed the local rate to roughly three times the global one. The other is the lag. Emissions already released have committed the system to further rise, and ice sheets answer over centuries rather than decades, so even a world that stopped emitting tomorrow inherits a rising sea for the rest of this century and beyond.

Add the atmosphere and the picture sharpens. Warmer air holds more water vapour, which means heavier rain when it comes, and warmer oceans feed stronger tropical cyclones that increasingly travel more slowly, dumping more on a given place over a longer period. Surge rides on top of a higher sea: a swell that once measured a metre and a half over a low atoll’s three-metre elevation now measures two metres over the same three, and the margin has gone. Come a storm at high tide with the surge behind it and the ocean crosses the island rather than stopping at the shore, taking topsoil, killing coconut palms, salting the pits where taro grows and filling the wells with brackish water. An island does not disappear in the storm. It comes back smaller, poorer and less able to feed the people on it.

That is how a country dies — slowly, unevenly, deniably. Nobody wakes to find their nation gone. They wake to find the taro has failed four years running, the well tastes of salt, the road floods at every spring tide, a third of the young men have entered a ballot for a visa they never expected to want, and the school has fourteen pupils where it used to have sixty.

Four Metres of Country

Nine islands and about twenty-six square kilometres of land make up Tuvalu, scattered across an exclusive economic zone of roughly nine hundred thousand square kilometres of ocean — the land smaller than a mid-sized English city, the sea it governs larger than Turkey.

Four and a half metres is the highest natural point in the country, on Funafuti. Most of the land sits between one and two. Nearly half the population crowds onto Fongafale, the main islet of the Funafuti atoll, which in places is barely wide enough to hold the runway bisecting it — a strip built by American engineers during the Second World War that serves as the country’s main public space and, in the drought of 2011, became the distribution point for water shipped in from New Zealand. At high tide the ocean no longer floods only over that runway; on the worst days it comes up through the ground beneath it.

What exposes Tuvalu is less the height of the land than the freshwater beneath it. An atoll is a ring of coral debris on a submerged volcano, and the only reason anything grows is that rain soaks through the porous ground and floats on the denser seawater below, forming a thin lens of fresh water that roots can reach. A metre or two deep at best, the lens sits in delicate balance with the sea. Raise the ocean a few centimetres and it thins and salinises; dig a borrow pit, as the engineers did for the runway, and it breaks there permanently; add a drought and the wells go salty within weeks. Three decades of measurement put the rise around Tuvalu at roughly fifteen centimetres, and the wells have registered it.

Pulaka, the giant swamp taro cultivated in dug pits for centuries, cannot tolerate salt. As the pits turn brackish the crop fails, and a food culture older than written history comes loose from its foundation. Imported rice and tinned meat take its place, raising the import bill, worsening the diet and further weakening the case for staying.

Sand, meanwhile, has become the newest thing on the islands. Around Funafuti, cutter suction dredges have been vacuuming the lagoon floor and pumping it into geotextile containers stacked higher and higher, building new ground above the forecast flood line — three phases, funded by Australia and the Green Climate Fund, and the largest construction project in the country’s history. An Australian coastal engineer working on it, James Lewis, reckons the programme has already enlarged Tuvalu’s landmass by more than a tenth. Maina Talia, the climate change minister, now stands on new land where the foreshore in front of his house used to be, and bristles at the phrase climate migrant. “Land in Tuvalu means life,” he says. “If you dismantle your relationship with the land in Tuvalu, you are no-one.”

Politics have tracked the water in the same way. In 2021 the foreign minister, Simon Kofe, recorded a speech for the Glasgow climate conference standing in a suit, knee-deep in the sea off Fongafale. A year later, at Sharm el-Sheikh, he announced that Tuvalu would build a digital replica of itself — a three-dimensional scan of every island, a digital identity system for a future diaspora, and an archive of songs, documents and ceremonies, so that a version of the country could keep holding elections and registering births when the physical version could not. Kofe also chaired the parliamentary committee behind the 2023 amendments, which fixed not only perpetual statehood but permanent maritime boundaries by coordinates and an archipelagic baseline, keeping the ocean legally attached to the state even if the land beneath the fishing grounds goes under.

Money is thinner than the politics. A substantial share of government revenue comes from leasing the internet domain .tv, which the streaming boom turned into a modest windfall — some ten million Australian dollars a year under the current arrangement with GoDaddy, trivial to a mid-sized economy and transformative to one with eleven thousand people. Fishing licences from the Parties to the Nauru Agreement add more, and the Tuvalu Trust Fund, established in 1987, has been managed to supplement the budget. Enough to buy time, not enough to pay for the coastal defence that holding the line indefinitely would require — and time is the commodity Tuvalu is short of.

Thirty-Three Pieces of Almost Nothing

Kiribati is harder still to hold in the mind. Thirty-three atolls and one raised coral island spread over three and a half million square kilometres of ocean, roughly the size of India, with a total land area smaller than Rome. Most of the population lives on South Tarawa, a strip a few hundred metres wide in places, carrying some sixty thousand people and almost no room to expand.

Banaba, the raised coral island that gives the country part of its name, climbs to eighty-one metres and is geologically a different thing from the atolls around it — with a different history to match. Its phosphate was mined to exhaustion by the British Phosphate Commission, and its people were moved, in stages and against their will, to Rabi Island in Fiji, where most of their descendants still live. Banaba could plausibly survive a very large rise in sea level. The Banabans are not on it.

Longer than almost anyone, this republic has been thinking about the problem. Two uninhabited islets, Tebua Tarawa and Abanuea, went under in 1999. In 2014, President Anote Tong completed the purchase of the Natoavatu estate on Vanua Levu in Fiji — roughly 2,220 hectares, about 5,500 acres, for around A$9.3 million — as a fallback homeland. Tong called the policy “migration with dignity”, a phrase that has aged well and badly at once: well, because it framed relocation as a choice rather than a defeat; badly, because more than a decade on the estate remains largely undeveloped and few i-Kiribati have moved to it. Most prefer to wait and see. The land is there. The people are elsewhere.

Tong’s successor, Taneti Maamau, shifted emphasis from buying escape to building walls, and under him the Temaiku reclamation on South Tarawa took shape: dredged sand, a raised platform, a new settlement designed to consolidate the population onto defensible ground. Sound logic, enormous cost for a country of Kiribati’s means, and an assumption that the thing being defended does not itself get overtopped.

Scale is where Kiribati differs from Tuvalu. Eleven thousand people can be relocated almost anywhere; a hundred and twenty thousand cannot, and the Fiji estate was bought with sixty to seventy thousand in mind. New Zealand’s Pacific Access Category admits a quota of i-Kiribati each year, alongside wider Pacific migration flows, and the country has been quietly exporting people for decades. That is a pressure valve and a slow national bleed together: atolls emptying of working-age adults, villages filling with the very old and the very young, remittances holding up the economy, and fewer people each year left to defend the land.

A Dome, a Lagoon and a Debt

No other nation in this discussion carries the Marshall Islands’ particular burden. Between 1946 and 1958 the United States detonated sixty-seven nuclear devices across Bikini and Enewetak atolls. Bikini’s people were moved, told it would be temporary, and have been moved repeatedly since; Enewetak’s returned to an atoll reorganised at the scale of a geological force.

Most visible among the residues is Runit Island: a crater left by one test, filled in the late 1970s with some 120,000 tonnes of radioactive soil and debris, capped with an eighteen-inch concrete dome roughly a hundred metres across. Nothing about it is sealed. The dome sits directly on porous coral in contact with the sea, and the cap has been cracking for years — fresh fissures were reported in March 2026 — while the surrounding land averages two metres of elevation, so the water table inside is tied to the ocean’s. As the sea rises, the water in the crater rises with it, and where the plutonium goes when it moves has no comfortable answer.

Majuro, the capital, is a thin, overdeveloped atoll carrying the seat of government and the bulk of the population; Ebeye, on Kwajalein, packs something like fifteen thousand people onto a third of a square kilometre, among the highest densities on earth. Unlike its neighbours, the Marshalls hold an advantage: the Compact of Free Association grants citizens the right to live and work in the United States, which means the country has been migrating quietly for decades and a diaspora of real size already exists in Arkansas, Hawaii and elsewhere. A lifeline, and a slow dissolution of the thing being preserved.

Washington holds full defence authority under the Compact, and the Marshallese have leveraged that relationship into climate diplomacy, arguing that the country whose tests made America’s Cold War possible should not now be left to drown. Freezing maritime boundaries has been another cause, and here the state has been among the loudest advocates — its economic zone vast, its land negligible, and a country with legally permanent waters retaining a future even without a territory.

Paradise Priced by the Square Metre

Most populous of the atoll states and most economically exposed, the Maldives has some 520,000 people on about 298 square kilometres of land across nearly twelve hundred islands, around two hundred of them inhabited. Natural elevation peaks at a shade over two metres. Malé, the capital, ranks among the densest cities on the planet, and an economy leaning on tourism means the assets most at risk — resorts, airports, beaches — are precisely the ones built at the waterline.

For decades now the Maldives has been building land, and building it deliberately high. Hulhumalé, reclaimed from a shallow lagoon beside the airport and settled from 2004, was constructed to sit roughly two metres above sea level, higher than Malé, with a designed capacity of a quarter of a million people. It is now a functioning city joined to the capital by bridge, and the clearest example anywhere of a country deciding that if the sea is coming, the ground will simply be built up to meet it.

Problems with the project are political and ecological rather than technical. Sand for the reclamation is dredged from the seabed, and dredging damages the reefs that shield the islands from waves; reclaimed land is dense, expensive and largely sold to the highest bidder, and critics argue the benefits accrue to developers and tourists rather than to fishing families whose access to the shore has been altered. Hulhumalé also answers only part of the question. It can absorb some of the population of Malé and the nearby islands. It cannot absorb the whole country, nor the loss of the reefs, which are the reason the Maldives exists as land at all.

Among gestures, the most famous remains the underwater cabinet meeting held by President Mohamed Nasheed in 2009, when ministers in scuba gear signed a document on the sea bed to dramatise the threat. No gesture could change the arithmetic. Buying land abroad — in India, Sri Lanka, Australia — has been considered and never quite committed to. Reclamation, defence, and the proposition that a country can be a set of islands in one decade and a set of fortified islands in the next, have taken its place.

An Island That Ate Its Own Past

Nauru is the odd one out, and instructively so — a raised coral island rather than an atoll, twenty-one square kilometres, with a central plateau climbing to seventy-one metres. Sea level rise alone cannot drown Nauru. What it can do is ruin the narrow coastal ring holding the airport, the port, the road and the entire built environment, while the phosphate-mined interior sits above it, pitted and useless.

That interior is the residue of the twentieth century’s most complete act of national self-consumption. Phosphate deposits made Nauru’s people, for a period in the 1970s, the highest per-capita income earners on earth; mining stripped the plateau to a moonscape of coral pinnacles, exported the soil, and left a country with money, no soil and nothing to grow. When the money was mismanaged and the deposits ran out, other revenue followed — a briefly notorious spell as a tax haven, the sale of fishing rights, and, more durably, the leasing of territory to Australia as an offshore processing centre for asylum seekers.

What Nauru demonstrates is a country hollowed out long before it is submerged. Habitability can be lost without land being lost, and that is the template for the other atoll states on the way down. Money can be earned, land bought, sovereignty maintained, and the country can still become a place from which everyone leaves.

Where the Hill Is Not Enough

It would be convenient to divide the Pacific into doomed atolls and safe high islands, and to console oneself that Pohnpei’s 791 metres, Vanuatu’s 1,877-metre Mount Tabwemasana, Samoa’s Mauga Silisili or Palau’s Mount Ngerchelchuus will keep their people dry. Convenient, and misleading.

High interiors come with low coastlines, and almost everything that matters sits on the coastline. Airports need flat land, and flat land on a volcanic island is coastal; ports need deep water, which means shore; towns grew where the boats landed. A country with an 800-metre mountain and a village at sea level has a village at sea level, and moving it uphill means moving onto slopes never farmed, never built on, and in many cases already occupied.

Nowhere is the point clearer than in the Federated States of Micronesia. Pohnpei and Kosrae are mountainous and reasonably secure, but the nation also includes the outlying atolls of Nukuoro, Kapingamarangi, Mokil and Pingelap — as exposed as Tuvalu, in a country whose capital sits on a mountain — along with the outer islands of Chuuk and Yap. A state can survive while parts of it cannot, and who pays to move a few hundred people from an atoll to a mountainous island in the same federation turns out to be surprisingly hard to answer. Most Pacific states have no legal mechanism, no dedicated fund and no agreed procedure for internal climate displacement, and in the smallest ones the arriving population can be a substantial fraction of the destination’s.

Reefs and marine protection have absorbed the attention of Palau, better resourced and better governed than most, which has become one of the loudest voices in the international legal campaign to force emissions cuts. Vanuatu, having lost a large share of its GDP to Cyclones Judy and Lola in 2023, led the campaign that produced the UN General Assembly resolution requesting an advisory opinion from the International Court of Justice on climate obligations — diplomacy that has shifted the legal terrain further than a decade of speeches. Both countries have the geography to survive. Neither can afford to be indifferent.

Solomon Islands is already producing corpses of a kind. A study published in 2016 in Environmental Research Letters, led by Simon Albert of the University of Queensland, used aerial and satellite imagery from 1947 to 2014 to examine thirty-three reef islands, and found five vegetated islands that had vanished entirely and six more in severe shoreline recession. Sea levels around the Solomons had risen at close to three times the global rate, making the archipelago a natural laboratory. Nuatambu, one of the eroded islands, lost eleven houses and half its habitable area in a handful of years. As a state, Solomon Islands is not going anywhere; seven hundred thousand people spread across nearly a thousand islands. Parts of it have already gone.

Salt in the Well

Dramatic as the country-level stories are, the mechanism of loss operates at a smaller scale, and it is worth describing plainly, because habitability collapses long before land disappears.

Start with water. No atoll has rivers or lakes; every drop of fresh water is rain that has soaked into porous ground and now floats on seawater. Thin at the best of times, the lens thickens after rain and thins in drought, holding a rough equilibrium with the sea that depends on the density difference between fresh and salt. Push the sea up and it is squeezed and thinned; over-pump it and salt water is drawn into the wells. Both are happening, and on the outer islands of Kiribati and the Marshall Islands the wells that go brackish are usually the first hard evidence anyone notices.

Soil comes next. Coral sand is not soil in any agricultural sense — almost no nutrients, almost no organic matter — and the fertility of an inhabited atoll comes from centuries of human labour: compost, fish scraps, coconut husks, taro pits dug down to the water table. Salt water kills the taro, salt spray kills breadfruit and banana, and coconut palms, the last to go, tolerate more but are not a farm once the understory has died.

Reefs go the same way. A healthy reef can cut wave energy reaching a shoreline by up to ninety-seven per cent, which is why an atoll with coral intact absorbs swells that would otherwise strip it clean. Warm water bleaches coral, and the events of 1998, 2010, 2016 and the record global bleaching that began in 2023 have killed enormous areas of reef; a bleached reef is not dead, and can recover within a decade if conditions allow, but recovery is now interrupted by the next event before the last has finished. Acidification — roughly thirty per cent more acidic than before the industrial revolution — slows the rate at which corals, molluscs and calcifying plankton build skeletons. The reef protecting the island is attacked from three directions at once.

Then there is the fish. Reef fish are the protein supply and the fallback when taro fails and wages do not come; warming water shifts the range of pelagic species, which matters a great deal for the offshore tuna fishery, and the offshore fishery is where the money is. Eight states, including Tuvalu, Kiribati, the Marshall Islands and Nauru, have collectively managed the world’s largest tuna resource with unexpected sophistication under the Parties to the Nauru Agreement, using a vessel day scheme to raise the price of access and generating around half a billion dollars a year in revenue for members. Remarkable for a group of microstates, and a resource whose location is set by oceanography rather than law, since a warming ocean moves fish.

And the graves. Cemeteries on an atoll are usually sited on the highest available ground, often a beach ridge, and they are the first things washed and the first dead displaced. Reburial is possible; the point is that a grandmother’s grave, on the ground where she was born, is not the same thing as a grave relocated to a suburb in Auckland.

A Politics of Lifeboats

Relocation is not a technical problem but a political one, and the political difficulty is that nobody wants to be the government that presided over the end of the country.

Of the arrangements attempted so far, the most substantial is the Falepili Union between Australia and Tuvalu, signed in November 2023 and in force since August 2024. It provides a pathway for up to 280 Tuvaluans a year to live, work and study in Australia with permanent residency, commits Australia to recognise Tuvalu’s statehood even if the islands are lost, and has been accompanied by more than A$110 million in Australian money for coastal adaptation and new land. The name is a Tuvaluan word for good neighbours.

Uptake outran every expectation. When the first ballot opened in 2025, somewhere between two-thirds and nine-tenths of the population applied — the ABC put the figure at 65 to 80 per cent, the United Nations at 90 — and the 280 visas were granted, with the first families arriving in Australia over the past year. Nobody involved had expected an application rate that high, and the awkward question it produced is now unavoidable. What happens to a country when most of its people hold a legal route out?

New Zealand’s Pacific Access Category does something similar with less ceremony, admitting a small annual quota from Tuvalu and Kiribati, and New Zealand’s wider Pacific migration policy has been the quiet backdrop to a great deal of Pacific life for decades. Right of residence in the United States comes to Marshallese, Palauans and Micronesians through the Compact of Free Association. Fiji, exposed itself but possessed of high ground and space, has been both destination and broker, having sold Kiribati the Vanua Levu estate and relocated several of its own villages, including Vunidogoloa in 2014 — generally cited as the first formally organised climate relocation of a village anywhere.

Bilateral, voluntary and small, every one of these arrangements leaves the same gaps. A country of eleven thousand can be absorbed by Australia without strain; a hundred and twenty thousand cannot be absorbed by New Zealand without a debate about housing, jobs and cultural accommodation that has not yet been held. Nobody has designed a system for the actual dissolution of a state, either: who inherits the seat at the United Nations, who takes over the fishing fleet, who holds the sovereign wealth fund, what happens to citizenship, and who is responsible for those who did not get a visa.

Criticism has come from inside, too. Enele Sopoaga, a former prime minister now in opposition, calls the treaty interventionist and imperialistic, and worries that the people leaving are precisely those between eighteen and forty-five — the active workforce of a country that needs one. Australia’s high commissioner in Funafuti insists the pathway runs both ways, and that the point is for Tuvaluans to come home with skills. Both claims can hold at once.

A darker possibility, discussed in the Pacific more openly than outsiders expect, is that the countries empty out before they flood. Migration is selective, and the young, educated and mobile leave first because they can, which raises the cost of maintaining the infrastructure that remains; a shrinking tax base and a growing stock of coastal defence needs make a fiscal vice. In theory the atolls could reach permanent submersion in the 2070s with only a few hundred people left to be rescued, having spent fifty years ceasing to function as a society. Legal state would survive. Lived country would have ended quietly.

A Country Without Ground

Because that outcome is visible on the horizon, a substantial body of international legal work has grown up around it, and in the past two years it has largely gone the islanders’ way.

Dating from 1933, the Montevideo Convention supplies the classic definition of a state: a permanent population, a defined territory, a government, and the capacity to enter into relations with other states. Comfortably satisfied are the first, third and fourth criteria. Territory is the problem. If land is submerged, is it still territory? Water does not leave a gap in the law where the land was; it covers the land, which remains the land, only wetter. The seabed goes nowhere. A submerged atoll is still a feature with coordinates, and coordinates can be written into a constitution.

Tuvalu has done exactly that. The 2023 amendments fixed the boundaries of the state and its maritime zones by reference to an archipelagic baseline and geographic coordinates, and declared them permanent regardless of what happens to the land. Several states have signed on to recognise a digital Tuvalu, and Australia’s treaty obligation makes recognition of a landless Tuvalu an explicit commitment rather than a courtesy. The Marshall Islands, Kiribati and the Maldives have pursued variations on the same theme, and the Rising Nations Initiative, launched by Pacific heads of state and hosted by the Global Centre for Climate Mobility, exists to push the question into the United Nations.

Then the courts and the UN’s law commission moved, in quick succession. In May 2024 the International Tribunal for the Law of the Sea held that greenhouse gas emissions constitute marine pollution and that states are obliged to protect the marine environment. The International Law Commission, after years on the question, concluded that the law of the sea does not require states to update baselines or outer maritime limits once these have been lawfully established, and set out practical tools — climate mobility agreements, humanitarian visas, safeguards against statelessness — for the people involved. On 23 July 2025, having heard nearly a hundred states and organisations argue the case in December 2024, with the Pacific among them, the International Court of Justice delivered its advisory opinion on the obligations of states in respect of climate change. Both bodies found strong support among states for the continuity of statehood. Both concluded that maritime zones do not retreat with the coastline.

Not binding, either of them, but close enough to shape behaviour. “We are not going to change the map for a crisis we did not cause,” the Marshall Islands president, Hilda Heine, told the UN’s high-level meeting on sea level rise. “The answer cannot be that our borders will be ambiguous or open for international plunder of our resources.”

Stakes here are financial as much as sentimental. An exclusive economic zone is not a gesture. The Marshall Islands governs two million square kilometres of ocean. Kiribati governs three and a half million. Tuvalu’s is roughly nine hundred thousand, and the tuna that swims through it is worth real money. If statehood collapsed with the land, the zones would become high seas, open to any fleet with a licence from nobody, and the Pacific microstates would lose the single most valuable asset they own. The legal campaign, in the end, is a campaign about fish.

Raising the Floor

Engineering has an answer, or several, and the answers are honest about their limits.

Reclamation comes first: build the land higher. The Maldives has done it at scale, the Netherlands has been doing it for a thousand years, Kiribati has planned it at Temaiku, and Tuvalu is now doing it around Funafuti with dredges and geotextile. The technique is straightforward — pump sand into a lagoon or behind a bund, compact it, build a sea wall, and set the finished ground two or three metres above the natural island. The result is defensible and can last a long time. Cost runs to tens of millions for a small island and billions for a national programme, and the dredging damages the reef that would otherwise have absorbed the waves.

Hard defence comes second: sea walls, revetments, breakwaters, groynes. The Maldives has walled much of Malé. Tuvalu has a seawall and breakwater programme funded largely by donors. Walls work best where there is money to maintain them, and they transfer the erosion problem along the coast; a wall on one island accelerates the loss of sand from its neighbour.

Nature offers a third route: reef restoration, coral nurseries, mangrove planting, beach nourishment. Mangroves absorb wave energy and accrete sediment, and a restored mangrove belt is cheaper and more self-maintaining than concrete. Coral restoration is slower and more fragile, and a nursery-grown coral is as vulnerable to the next bleaching event as a wild one. Reef restoration cannot outrun warming; it can only buy time.

Most speculative, and most interesting, is the fourth: floating and amphibious settlement. The Maldives has partnered with a Dutch firm to build a floating city of modular platforms in a lagoon near Malé, designed to house around twenty thousand people, moored rather than reclaimed, rising and falling with the water. UN-Habitat has promoted the concept through a prototype project in Busan, South Korea. Floating cities solve the flooding problem by removing the ground entirely, and they create a new one in that they need anchoring, utilities, sewage, and a legal regime for a settlement that is not quite on land. Whether they become a model or a curiosity depends on costs not yet demonstrated at scale.

None of these options is a substitute for emissions reduction, and all of them are expensive. The honest engineering assessment is that money can hold a small atoll above water for a long time, that the limit is set by how much a country is willing to spend and how many countries are willing to help, and that no amount of engineering restores the freshwater lens, the taro pits or the reef fishery that made the place worth defending.

What a Vanishing Economy Costs

Economics is where Pacific climate diplomacy becomes brutally concrete, and where the difference between microstates and larger coastal economies is sharpest.

Microstate economies are small and narrow. Tuvalu’s government revenue leans on the .tv domain lease, fishing licences and donor support, with a sovereign trust fund providing a buffer; Kiribati’s rests on fishing licences and foreign aid. The Marshall Islands depends on Compact payments and fisheries, Nauru on fisheries, the trust fund and the Australian detention centre. None of these countries can self-finance the coastal defence they need, and all of them are, in effect, negotiating for their survival with the same limited diplomatic tools: moral authority, legal innovation, and the leverage of votes and ocean.

Leverage of that kind is not trivial. The Pacific Small Island Developing States bloc has been unusually effective at the United Nations, and the Loss and Damage Fund agreed at the Sharm el-Sheikh conference in 2022 and operationalised in Dubai in 2023 was substantially a Pacific achievement, with pledges approaching a billion dollars. The gap between pledges and need is enormous. The Global Centre on Adaptation has estimated that the adaptation needs of small island states run into the hundreds of billions over the coming decades, and the fund’s initial capitalisation is a rounding error against the number.

For the larger coastal economies the calculus is different and larger. Bangladesh loses an estimated billion dollars a year to cyclones, roughly seven-tenths of a per cent of GDP, and the figure rises with each storm season. Egypt’s Nile Delta carries a large share of national industry and agriculture, and Alexandria is a city of five million that sits on subsiding ground. Vietnam’s Mekong Delta produces over half the country’s rice and hosts seventeen to twenty-one million people on land that is on average less than a metre above sea level. Guyana’s coast is the entire national economy, protected by a seawall and a canal system built by the Dutch and maintained ever since, with the capital sitting six feet below high tide. The Netherlands spends a substantial share of its national income on water management and has the GDP to do so. Guyana does not.

Asymmetry is the point. The countries with the least capacity to pay for defence are the ones with the most exposure, and the countries with the most exposure often bear the least responsibility for the emissions causing it. Pacific states contribute a negligible fraction of global greenhouse gases. Their argument, made repeatedly by António Guterres among others, is that this gives them a moral authority to demand action from the emitters. The emitters have, so far, acknowledged the argument and not matched it with money.

Memory, and Its Erasure

A category of loss that does not appear in the economics and only occasionally in the law is the one islanders themselves raise most insistently.

An atoll culture is a maritime culture, and much of what makes it distinct is tied to a specific reef, a passage, a fishing ground. Navigation knowledge, in the Carolinian and Marshallese traditions, is encoded in stick charts and star compasses and in the memory of swell patterns, and the memory is anchored to places that will be underwater. A navigator who can find an island by reading the way the ocean moves across a reef holds a map of the sea floor, and the sea floor does not move, which is a small consolation.

Language is a second casualty, and already in retreat. Tuvaluan, Kiribati, Marshallese, Nauruan and the languages of the outer atolls are all small, and migration to English-speaking destinations accelerates the shift. A community of eleven thousand speakers scattered across three continents by visa ballots does not hold its language the way a settled community does. Tuvalu’s answer, in part, is the archive: the recorded songs, the scanned islands, the digital identity that lets a child in Melbourne still belong to a country in the Pacific.

Ritual is the third. Funerals, weddings, the first-fish ceremonies, the fatele danced in the meeting house, the community work on the seawall — these assume a village, a church and a shoreline. A dispersed community can hold the practices, but it holds them differently, and in the second and third generation the difference becomes the substance. Pacific states know this, which is why so much of their diplomacy focuses on maintaining the entity rather than merely saving the people. A country that keeps its seat at the UN, its EEZ, its constitution and its coordinates can plausibly argue that it still exists, and the argument is not only for the benefit of foreigners.

Next in Line

Everything described so far concerns a few hundred thousand people. Tens of millions are next, and the discussion becomes uncomfortably large.

Bangladesh is the largest single exposure. Around twenty million people live in the coastal belt, two-thirds of the country sits less than five metres above sea level, and a one-metre rise would inundate a sixth of the land area and displace millions. The country has adapted with a stubbornness that deserves more credit than it gets: cyclone shelters, early warning, floating gardens, salt-tolerant rice, the world’s largest mangrove restoration programme in the Sundarbans. Adaptation has already cut cyclone deaths from the hundreds of thousands of the 1970s to the low tens. It has not stopped the salinisation of drinking water, the loss of farmland, or the migration of rural families into the slums of Dhaka and Chittagong. Bangladesh’s real problem is that adaptation is a treadmill, and the belt is running faster.

Vietnam’s Mekong Delta is the second. Seventeen to twenty-one million people, over half the national rice crop, a major share of aquaculture and a substantial part of the export economy sit on a delta whose average elevation is under a metre, and which is subsiding under its own weight as groundwater is pumped for shrimp farms and cities. Dams upstream in China and Laos have cut the sediment that used to rebuild the delta each flood season, so the land is sinking while the sea is rising, and the two together are far more than either alone. Several studies project that much or most of the delta will fall below sea level this century. Vietnamese engineers are building sluice gates, raising roads and experimenting with controlled flooding, but the delta is not a small problem with a small solution.

Egypt’s Nile Delta is third, and its situation has an ancient cause. The Aswan High Dam, completed in 1970, ended the annual flood that deposited fresh silt across the delta, and the delta has been compacting ever since. Alexandria subsides by a few millimetres a year on its own, and the surrounding ground subsides faster. Sea level rise is added on top. The city, home to five million people and a large share of Egypt’s industry, sits on the Mediterranean behind a corniche and a breakwater, with nowhere to retreat to because the ground behind it is also low. UNESCO recognised Alexandria as Africa’s first tsunami-ready community in 2024, which is a real achievement and also an admission of how exposed it is.

Indonesia’s problem is the sharpest of all, and the sharpest lesson. North Jakarta has sunk more than two and a half metres in a decade in places, at up to twenty-five centimetres a year, because industry and residents pumped groundwater after the piped network failed to keep up with the city’s growth. The subsidence is a self-inflicted wound that compounds sea level rise, and the response has been dramatic: Indonesia has moved its capital to Nusantara in Borneo, which became the country’s official seat of government in 2025 at a cost running into tens of billions of dollars. Moving a capital is the largest adaptation measure any country has taken, and it is worth noticing that it does not move the twelve million people of Jakarta.

Lagos and the Niger Delta come next, where subsidence, surge and erosion meet an exploding population; Bangkok, sinking at one to two centimetres a year behind a rising Gulf; Guyana and Suriname, where the entire organised economy sits on a narrow, artificially drained coastal plain; the Bahamas and the Caribbean, where hurricanes now arrive with a violence the building stock was not designed for; Louisiana and the American Gulf and Atlantic coasts, where subsidence and surge have already produced the first government-funded village relocations in the continental United States. And the Netherlands, the exception that proves the rule: a quarter of its land below sea level, a decision to remain, and a willingness to pay.

Four Dates

It helps to lay the sequence out, because the order of events is not the order in which they are usually discussed.

By 2030, the changes are still incremental and mostly absorbable. King tides reach further, flooding becomes routine in a handful of atoll capitals, the first wells turn permanently brackish, and the first relocation programmes move from planning to construction. Sea level rise by mid-century is largely locked in regardless of emissions, somewhere between fifteen and thirty centimetres depending on scenario, and this is the part of the future no policy can alter.

By 2050, the arithmetic bites. Several atolls will need continuous defence to remain habitable, and the cost of that defence will exceed the GDP of the countries maintaining it. A substantial share of the population of Tuvalu and Kiribati will have migrated, most of them to Australia and New Zealand. The Mekong Delta and the Nile Delta will be visibly losing land. Jakarta’s northern districts will be dependent on pumps and a sea wall. The once-in-a-century flood will be arriving roughly annually in much of the tropical Pacific, and coral reefs will have passed through at least two more global bleaching events.

By 2070, the first atolls become candidates for abandonment, and the legal machinery built over the previous fifty years gets its test. Some countries will retain populations on defended or reclaimed ground; others will become states whose citizens live elsewhere and whose territory is a lagoon with coordinates. Delta cities will be spending a share of national wealth on water, and the terms of trade between adaptation and retreat will have shifted in favour of retreat.

By 2100, the global figure lands somewhere between half a metre and a metre under most scenarios, with more than a metre at the high end and considerably more beyond the century. Combined with subsidence, storm surge and salt intrusion, this is enough to remove the habitability of a great deal of very expensive and very populated land. The question by then is no longer which countries are at risk. It is which countries still have functioning coastlines, and how many of them could pay for the difference.

What Might Hold

It is worth resisting the temptation to write the Pacific off, because the region has a habit of confounding the expectations of outsiders.

Coral reefs can recover. After the 1998 bleaching event, which killed a large share of corals in the western Indian Ocean, some reefs regained much of their cover within a decade. Hold warming near the lower end of the range and Pacific reefs could persist in a changed form, and a living reef is a functioning breakwater. Mangrove restoration is faster, cheaper and has a demonstrable record. Seagrass and reef flat restoration improve sediment accretion, and there is evidence that a well-managed reef flat can raise the sea floor slightly over time.

Defended and reclaimed ground can outlast the century; the Dutch have proved that a determined country with money can live below sea level indefinitely, and a partial version of their model is within reach of a raised, walled, pumped island with a few thousand people and a functioning economy.

Legal innovations have held better than anyone expected. Perpetual statehood, permanent maritime boundaries and recognised digital territory reinforce each other, and the trend in international law is running in the Pacific’s favour. With the International Court of Justice and the law of the sea now on record, a submerged atoll state with a fixed EEZ has moved from thought experiment to live legal question, and a country with an EEZ, a seat at the UN and a diaspora can be argued into existence.

What will not hold is the assumption that the people can stay. Every projection that matters includes migration, and the honest conclusion of every serious study of the small island states is that adaptation and migration are not alternatives. They are the same policy, and the question is only how the two are sequenced, funded and dignified.

A Century Hence

There is a version of this story told so often it has become almost comfortable: the Pacific islands as the world’s conscience, the canaries in the coal mine, the small nations whose sacrifice will teach the large ones a lesson before it is too late for the large ones. True, and flattering to the large nations, because it implies that the small ones are the ones being tested.

They are not the ones being tested. They are the ones being removed.

What is happening to Tuvalu, Kiribati, the Marshall Islands and the Maldives is not a warning but an outcome, arriving early, in countries that contributed almost nothing to the problem and have been the most articulate in describing it. Applause has been given, repeatedly, to people standing in the water in their best suits, and applause has not been converted into emissions reduction or money — and will not be, because the coordination it requires across a hundred and ninety countries has never yet been achieved on any subject.

So the small states have done what small states do. Law has been adapted, land bought abroad, ground built where there was none, visas negotiated for their own citizens, islands scanned into servers, and a claim written into constitutions that they will continue to exist when the ground is gone. Something admirable in the thoroughness, something grim in the necessity.

In fifty years, the map will still show Tuvalu — defended by a sea wall, perhaps, or with most of its citizens in Australia, or with a functioning parliament in a digital space and a lagoon where the capital used to be. The country will not have been erased, because its people will have refused to allow it. But the thing that made the country a place — the taro pit, the well, the church on the beach ridge, the grandmother’s grave, the reef where the fishing is good — will have been substantially surrendered. And the same forces that did that to a country of eleven thousand will, over the following decades, do something comparable to the deltas where hundreds of millions live.

The ledger is not yet closed, and the entries are not yet fixed. What is fixed is the direction of the water.

A Note on Sources

Global and regional sea level projections come from the Intergovernmental Panel on Climate Change’s Sixth Assessment Report, Working Group I, Chapter 9, and the Synthesis Report, Figure 3.4, along with NASA’s sea level projection tool built on the same data; the World Meteorological Organization’s floor and tail figures for 2100 appear in UN document A/81/74. The western Pacific rate of rise, at roughly two to three times the global average, is documented by the US Geological Survey and by the WMO’s Pacific climate reports. Pacific atoll elevation figures come from USGS atoll studies and national statistics.

The 2016 Solomon Islands study is Albert and colleagues, “Interactions between sea-level rise and wave exposure on reef island dynamics in the Solomon Islands,” Environmental Research Letters 11(5), summarised contemporaneously by the BBC and New Scientist. Marshall Islands nuclear and Runit Dome material draws on reporting by the Los Angeles Times and ABC News, on the March 2026 reporting of fresh cracking, and on US Department of Energy programme documentation. Kiribati’s 2014 purchase of the Natoavatu estate in Fiji was reported by The Guardian and the Pacific Islands News Association. Tuvalu’s constitutional amendment is documented by the Library of Congress Global Legal Monitor and analysed by the Development Policy Centre at the Australian National University. Tuvalu’s digital nation programme is described by UN News and by the government’s own material. Falepili Union detail, the 2025 ballot, the A$110 million figure, the Coastal Adaptation Project and the quotation from Maina Talia come from ABC News reporting of May 2026; the 90 per cent application figure and the quotation from Hilda Heine come from UN News, September 2026. Hulhumalé’s elevation and capacity are documented by the Global Center on Adaptation and in the academic literature on the Maldives’ reclamation programme.

Carteret Islands relocation is documented by The New Humanitarian and UNICEF. Jakarta’s subsidence rates and the Nusantara project are covered by the BBC and Indonesian and international reporting. Mekong Delta projections come from Nature Communications and Utrecht University’s delta research programme. Alexandria’s subsidence and the Nile Delta’s exposure are documented by The Guardian, the Geological Society of America’s GSA Today, and the UNDP’s Egyptian coastal adaptation programme. Guyana’s coastal vulnerability is drawn from CARICOM statements and Guyanese press reporting. Netherlands figures are from the IMF’s 2024 climate risk technical note and Dutch water sector documentation. Tuna revenue figures come from the Pacific Islands Forum Fisheries Agency and published analyses of the Parties to the Nauru Agreement’s vessel day scheme. The .tv domain revenue figures are reported by ABC Pacific and PRX’s The World.

Legal material draws on the ICJ case file for Obligations of States in respect of Climate Change (advisory opinion, 23 July 2025), the International Tribunal for the Law of the Sea’s 2024 advisory opinion, the International Law Commission’s work on sea-level rise in relation to international law, and UN News’s September 2026 summary of both.