The Slow-Motion Displacement of the Pacific

1Background: a hazard with no event

Disaster science is organized around events: a cyclone makes landfall, a river crests, an earthquake strikes. Sea-level rise on low-lying islands does not fit that template. It raises the baseline against which every other hazard operates, so that a king tide or a swell that once caused a nuisance begins to cause damage, and the damage accumulates without a date that anyone can point to.

The exposure is structural. Tuvalu’s mean elevation is about 2 m, and more than half of its people live on a single islet.^[7] More broadly, the IPCC estimates that about half of the Pacific’s population lives within 10 km of the coast, with at least half of its infrastructure within 500 m.^[3]

This brief asks three questions. What does the physical evidence show? Through what mechanisms do islands become hard to live on? And what are governments and legal institutions doing in response?

2The physical evidence

Global mean sea level rose by 0.20 m (very likely range 0.15 to 0.25 m) between 1901 and 2018. The rate has increased through the period, from 1.3 mm a year in 1901–1971 to 1.9 mm in 1971–2006 and 3.7 mm in 2006–2018, and human influence was very likely the main driver of the increase since at least 1971 (Figure 1).^[1]

Rates in the Pacific differ from the global mean. One analysis of Tuvalu estimated local rise at about 3.9 ± 0.4 mm a year over four decades, twice the global average for that period.^[2] NASA reports that sea level there is nearly 15 cm higher than it was 30 years ago, rising at about 5 mm a year, roughly 1.5 times the global rate, and expects that rate to more than double by 2100.^[7] The two estimates cover different periods and are not directly comparable, but both place local rise above the global rate.

Projections to 2100 depend on emissions. Relative to 1995–2014, the IPCC’s likely range for global rise is 0.28–0.55 m under its very low emissions scenario (SSP1-1.9) and 0.63–1.01 m under its very high scenario (SSP5-8.5). A rise approaching 2 m by 2100 under the highest scenario cannot be ruled out because of deep uncertainty in ice-sheet processes, although confidence in that outcome is low. Sea level is also committed to keep rising for centuries to millennia.^[1]

3Land is not the same as habitability

The popular image of Pacific atolls is of land steadily disappearing. The best documented case complicates it. Using remotely sensed data for all 101 islands of Tuvalu, Kench and colleagues found a net increase in land area of 73.5 ha (2.9%) over four decades, with land area increasing in eight of nine atolls. Change was uneven: 74% of islands grew and 27% shrank (percentages as reported).^[2] The authors concluded that islands behave as dynamic features and will persist as places to live over the next century, which widens the range of adaptation options.^[2]

Elsewhere the evidence is starker. In the Solomon Islands, an analysis of 33 reef islands between 1947 and 2014 identified five vegetated islands that had vanished and six more with severe shoreline recession. Erosion destroyed two village sites that had existed since at least 1935 and forced relocations.^[8] Recession was substantially higher where wave energy was high, indicating a synergy between sea-level rise and waves. The authors also note that earlier research in the western Pacific had attributed most shoreline change to extreme events, seawalls and inappropriate development rather than sea-level rise alone, a reminder that local drivers matter.^[8]

These findings are compatible rather than contradictory, because land area is a poor proxy for habitability. The IPCC assesses with high confidence that reef island and coastal habitability is expected to decrease because of higher temperatures, extreme sea levels and the degradation of buffering ecosystems, and that reduced habitability is an overarching risk even at 1.5°C of global warming.^[3] In its answer to the question of whether some islands could become uninhabitable, the chapter describes low-lying islands that may flood at every high tide or during storms, with freshwater supplies and soils repeatedly contaminated by salt, and concludes that it is unlikely such locations would remain habitable unless emissions are reduced or adaptation acceptable to the population is found.^[3]

4How habitability erodes

Three channels matter most, and they compound one another.

Flooding. The IPCC expects that an additional 5–10 cm of sea-level rise, anticipated for roughly 2030–2050, will double flooding frequency across much of the tropical Pacific and Indian Ocean. It also finds that some Pacific atoll islands will likely experience annual wave-driven flooding over their entire surface from the 2060s–2070s to the 2090s under a high emissions pathway, although future reef growth may delay the onset (limited evidence, low agreement).^[3]

Freshwater. Freshwater is the quieter constraint. The IPCC notes that freshwater resources in some Pacific atolls could be significantly affected by 0.40 m of sea-level rise. As an indication of magnitude, an 11–36% reduction in the volume of the freshwater lens is estimated for small atoll islands in the Maldives (area under 0.6 km²).^[3] Where flooding repeatedly salts soils and water, the consequences cascade into food supply, ecosystems and economic activity.^[3]

Livelihoods. The vulnerability of communities that rely on coral reef systems may exceed adaptation limits well before 2100 even under a low emissions pathway (high confidence), and most Pacific Island Countries could see declines of 50% or more in maximum fish catch potential by 2100 relative to 1980–2000 under both a low and a high emissions scenario.^[3]

Gradual change also compounds sudden shocks. Destruction from Tropical Cyclone Winston in 2016 exceeded 20% of Fiji’s GDP,^[3] and each such event lands on a population whose margins of safety are already narrowing.

5Mobility, law and statehood

As habitability narrows, the question shifts from whether people will move to the terms on which they can. Four instruments show how quickly the legal landscape is changing (Table 1).

The first international test came in 2020, when the UN Human Rights Committee considered the complaint of Ioane Teitiota, a Kiribati national who argued that New Zealand’s refusal of protection violated his right to life. The Committee dismissed the complaint on the merits, but observed that, given the extreme risk of an entire country becoming submerged, the conditions of life in such a country “may become incompatible with the right to life with dignity.”^[10] The UN human rights office framed the decision as opening the door to future climate-related asylum claims.^[11]

In 2023, Pacific Islands Forum Leaders declared that international law presumes the continuity of statehood, that the statehood and sovereignty of Forum members will continue “notwithstanding the impact of climate change-related sea-level rise,” and that members bear responsibility for protecting affected people, including their human rights, culture and dignity.^[4]

The most concrete step is bilateral. The Falepili Union entered into force on 28 August 2024. Beyond recognizing Tuvalu’s continuing statehood, it creates a pathway for up to 280 Tuvaluan citizens a year to live, work or study in Australia, and obliges Australia to respond when Tuvalu requests help after a major natural disaster.^[5] Demand has been intense: the first ballot drew 8,750 registrations, a figure that includes family members, equivalent to 82% of the country’s 10,643 residents at the 2022 census.^[12] Registration is not departure, but the number indicates how widely the option is valued.

Forecasting how many people will ultimately move remains difficult. The IPCC expects climate-related migration to increase, but finds that its drivers and outcomes are highly context-specific and that insufficient evidence exists to estimate numbers of climate-related migrants now or in the future (medium evidence, high agreement).^[3]

6Adaptation in place, and its limits

Adaptation in place remains the first line of response, and Tuvalu’s program is among the most ambitious. The Tuvalu Coastal Adaptation Project, supported by US$36 million from the Green Climate Fund and US$2.9 million from the government, is designed to protect 2,780 m of high-value vulnerable coastline and to build national capacity for coastal management.^[13] UN News reported in April 2026 that the project is creating more than seven hectares of new land across three islands, with a further eight hectares planned along the southern shoreline of Funafuti, at a cost of close to US$55 million so far.^[14]

Such programs are exceptional. The IPCC finds that for many small islands adaptation is incremental and does not match the scale of extreme or compounding events (high confidence), that much of it remains small in scale and sectoral, and that small islands face challenges in accessing adaptation finance.^[3]

Where adaptation in place is not enough, relocation becomes the option of last resort, a framing that Fiji’s own procedures adopt explicitly. Fiji endorsed its Planned Relocation Guidelines in 2018^[15] and in 2019 enacted a statutory trust fund to finance relocation.^[9] The experience of Vunidogoloa, relocated in 2014 to a site about 1 km inland on the same clan land, shows both the promise and the cost. A longitudinal study of the village, with 26 households and about 140 residents, found reduced exposure to coastal flooding and storm surge, fewer water-borne diseases and better access to services. It also found lost access to fishing grounds, livelihood projects that did not produce a sustained transition, incomplete housing and infrastructure, and new exposure to cyclone winds and heat.^[6]

7Implications for disaster risk management

Analysis. This section is Disaster Science’s interpretation of the evidence above. It is not drawn from the cited sources.

Treat habitability as the risk variable. Event-based metrics, such as losses per cyclone, miss gradual change. Indicators of habitability, including flood-day frequency, freshwater availability and the share of livelihoods that depend on reefs, give planners an earlier signal and a defensible trigger for decisions.

Plan mobility before the crisis, and finance it like mitigation. Fiji’s trust fund and the Falepili Union both create institutions before a mass displacement event rather than after one. The IPCC’s observation that institutional and legal systems are often inadequately prepared for large-scale settlement relocation^[3] suggests that preparation is the scarce resource.

Protect the natural infrastructure that buys time. The IPCC notes that intact coral reefs, seagrass meadows and mangroves provide ecosystem services important to island communities, that their benefits cannot easily be replaced if they are lost, and that coastal flooding will intensify unless these ecosystems can keep pace with sea-level rise through vertical growth.^[3] Nature based measures are therefore not an alternative to hard protection or mobility. They are the layer that determines how much time the other two have.

Design for both outcomes. Because the Tuvalu evidence shows islands persisting physically while habitability declines, planning should avoid two errors: abandoning places that can still be made habitable, and defending places that cannot. A framework that states its thresholds in advance, such as the freshwater and flooding conditions under which a community would choose to move, serves people better than one that discovers them during an emergency.

8Uncertainty and limitations

Projections. Sea-level ranges reflect both emissions scenarios and deep uncertainty in ice-sheet processes; the IPCC treats a rise approaching 2 m by 2100 under the highest scenario as possible but assesses confidence in it as low.^[1]

Island scale. The Tuvalu result should not be generalized to every island. The IPCC rates the evidence for increased reef island erosion as limited with medium agreement, and the evidence for the timing of annual wave-driven flooding as limited with low agreement.^[3]

Evidence on losses. The IPCC reports that methods to assess climate-induced loss and damage remain largely undeveloped for small islands, and that there are no robust methodologies to infer attribution.^[3]

Scope. This brief synthesizes published assessments and peer-reviewed studies. It presents no original field data and is an editorial synthesis that has not undergone peer review.