Pacific Dataviz Challenge 2026 · Climate change · Interactive Dataviz

Vanishing
Ground.

Sea-level rise driven by climate change is already erasing Pacific reef islands — not as a future risk, but as a measured, ongoing loss.

Nuatambu, Sogomou, and Hetaheta's shapes below are the actual satellite-traced coastline for that island in that year (Digital Earth Pacific, 1999–2023). Kale — the one Isabel island in this set that vanished completely — gets its own timeline further down, built from the source paper's own figure rather than satellite imagery.

Solomon Islands sea level: ~3 mm/yr since 1950, 7–10 mm/yr since 1994 — among the fastest rises on Earth
5
Isabel reef islands already gone entirely (1962–2002)
4
more tracked below (including Kale, now fully gone)
25
years of real satellite-traced shoreline, not projection
1999
1999–2023: real Landsat-derived shoreline years for Nuatambu, Sogomou & Hetaheta
19992005201120172023
Kale Island — its own timeline, its own source
1947
1947–2014: Albert et al.'s own traced outlines (fig 2c) at 1947/1962/2002, then the 2011 remnant and full 2014 loss — Kale was already too small to satellite-trace before the Landsat record in the section above even begins
19471962200220112014

This is a real place, not just a data point

Kale is the ancestral land of Gladys Habu Bartlett's grandparents — a Solomon Islands pharmacist, climate advocate, and the country's former Miss Solomon Islands (2019/20). At fourteen she began photographing the island on family visits; within a few years, scientists confirmed it had fully submerged.

In 2026, ten years after that confirmation, British sculptor Jason deCaires Taylor installed an underwater memorial in the waters where Kale once stood — a life-size figure modelled on Gladys, unveiled together with the island's landowners and Isabel Province's community. The sculpture is inscribed with the island's own timeline — 2006, when tides first reached the land; 2016, when it was confirmed gone — alongside projected dates for 2036 and 2046, marking how far the same rise is expected to continue.

Elsewhere in the Pacific — documented, but outside this study's imagery
Why this is happening, and what can be done

The islands above aren't eroding for one single reason, and the response to that erosion isn't simple either. Three things worth understanding: what's actually driving the loss, what it costs the people who live with it, and which responses have — and haven't — worked elsewhere in the Pacific.

What's driving it

Solomon Islands sea levels are rising at nearly three times the global average — a documented regional "hotspot," part natural climate variability, part the same warming driving sea-level rise worldwide.

But sea-level rise alone doesn't fully explain the pattern. Albert et al.'s own study found a synergistic interaction with wave energy: islands exposed to high wave energy eroded dramatically, while a dozen similar islands in a sheltered, low-wave-energy part of the same province showed almost no change over the same period, despite facing the same rising sea. Exposure to waves appears to be what turns background sea-level rise into visible land loss.

It's also worth being precise about scope: other studies of reef islands elsewhere in the western Pacific have found that most shoreline change to date comes from storms, poorly-sited seawalls, and coastal development — not sea-level rise in isolation. The clear sea-level signal documented here is specific to this wave-exposed part of Isabel and Choiseul, not a claim that every eroding Pacific shoreline has the same cause.

What it costs people

Two villages shown on this map — inhabited since at least 1935 — have already been destroyed by shoreline recession and relocated. Nuatambu's families rebuilt without outside support, a pattern repeated at similarly small, remote communities nearby.

Choiseul's provincial capital, Taro, is set to become the first provincial capital anywhere in the Pacific to relocate its entire population and government services — a planned, internationally-funded move to the mainland, in contrast to the unplanned relocations smaller communities have had to manage alone.

Beyond lost land, communities elsewhere in Solomon Islands report saltwater reaching garden plots and swamp taro crops — a staple food — failing as a direct result. Customary land tenure gives displaced families a place to resettle on ancestral land, but it also means close-knit villages can fragment into small, scattered hamlets once they disperse to different family land.

What can be done

Concrete seawalls are the most visible response, but their track record is mixed. Case studies from Fiji found seawalls can trap water on the landward side, worsening the flooding they were built to prevent — a documented pattern researchers call maladaptation, and one already showing up in Solomon Islands: a seawall built for one community in 2021 was reported rusted and failing within two years.

Nature-based approaches — replanting mangroves, protecting reef structure — cost less, avoid that failure mode, and are increasingly paired with engineered structures in "hybrid" designs rather than used alone, though they take longer to establish and offer less immediate protection during that time.

Planned relocation, started early and backed by resources, preserves communities far better than the unplanned, last-resort moves smaller islands have had to make alone. And the same 2016 study behind this whole visualization makes its own recommendation directly: pairing local and traditional knowledge with technical scientific assessment, rather than relying on either alone, is what actually guides good adaptation decisions.

Sources for this interactive Dataviz

  1. Digital Earth Pacific, Coastlines product (dep_ls_coastlines) — Landsat-derived annual shoreline positions, 1999–2023, retrieved directly from the public dataset (s3://dep-public-data/dep_ls_coastlines/). Drives the live shape for Nuatambu, Sogomou, and Hetaheta.
  2. Albert, S., Leon, J., Grinham, A., Church, J., Gibbes, B. & Woodroffe, C. (2016), Interactions between sea-level rise and wave exposure on reef island dynamics in the Solomon Islands, Environmental Research Letters 11(5) 054011 — open access. Table 1 supplies the measured area figures for all four islands and the exact areas Kale's traced shape is rescaled to; Figure 2(c) supplies Kale's traced outlines directly. The paper's text supplies the sea-level figures (3 mm/yr since 1950, 7–10 mm/yr since 1994, citing Becker et al. 2012).
  3. Kiribati Climate Change Office, case study on Abaiang Atoll (climate.gov.ki) — basis for the Tebunginako figures.
  4. Reporting on Funafuti Atoll flood projections and Tuvalu's 2022 land-reclamation adaptation plan.
  5. Reporting on Taro's planned relocation as Choiseul Province's capital (ABC News, Scientific American, Climate Change News) and on customary land tenure resettlement patterns in Solomon Islands (The Conversation, 2016).
  6. Human Rights Watch (2025), "There's Just No More Land": Community-led Planned Relocation as Last-resort Adaptation to Sea Level Rise in Solomon Islands — basis for the Walande food security and seawall-failure account.
  7. Piggott-McKellar et al., reporting on seawall maladaptation case studies in Vanua Levu, Fiji; Global Center on Adaptation, Solutions for Vanishing Coastlines: Adapting to Coastal Erosion in Small Island Developing States — basis for the mitigation comparison.
  8. ABC Pacific, Gladys Habu once picnicked on Kale, a pristine tribal island. By 2014, it was gone (2024) and 'Come see the urgency': Solomon Siren marks 10 years since the loss of Kale Island (2026); Smithsonian Magazine and Solomon Star News reporting on the Solomon Siren sculpture and Gladys Habu Bartlett's advocacy; GOV.UK, Commonwealth Points of Light award citation (2021) — basis for the Kale/Gladys Habu Bartlett memorial section.

Reconstructing the satellite traces. Nuatambu, Sogomou, and Hetaheta's shapes come from Digital Earth Pacific's raw shoreline data, which stores each year's coastline as two or more disconnected line fragments rather than one closed loop — a tiling artifact of the source product, not missing coastline. Each year's fragments are chained back into a complete loop by matching their endpoints. Hetaheta's traced coastline also includes a genuinely detached eastern remnant in 2010–2012 (under 800 m², gone by 2013), shown highlighted; Sogomou's satellite record showed only isolated single-year fragments (2010, 2022) rather than a sustained pattern, so those are left out rather than presented as a real trend.

Why Kale has its own timeline. Kale's documented history (1947–2014) mostly predates the 1999–2023 satellite record used for the other three islands, and by the time its remnant was small enough for the story to matter, it was also too small (well under one 30m Landsat pixel) for reliable satellite tracing. Its four shapes (1947, 1962, 2002, and a small 2011 remnant) are instead hand-digitised directly from Albert et al. 2016's own Figure 2(c), extracted by exact pixel colour and rescaled to match that paper's measured areas precisely. Kale gets its own independent 1947–2014 timeline rather than sharing the main slider, so its full range — including the dramatic 1947 shape — stays reachable. Years between the four traced outlines show a smooth estimate interpolated between the nearest real tracings, labelled as such in the caption; only 2014 has no traced outline at all, matching the source figure's own caption that Kale "was completely displaced by 2014."

A note on rigour. Several details here were wrong before they were right — a self-intersection check with a floating-point bug once justified over-smoothing Sogomou's coastline, and an over-tight merge tolerance once misclassified one of Hetaheta's ordinary single-loop years as a detached island. Both were caught by cross-checking derived numbers against independent sources (Albert et al.'s measured areas, the raw fragment gap distances) rather than trusting the first plausible-looking result.