Tuvalu's National Digital Twin — AI-Assisted Mapping of a Vanishing Coastline
Tuvalu
Tuvalu's Department of Lands and Survey built a machine-learning-assisted digital twin of Funafuti under the Pacific Community's Digital Earth Pacific …
Fiji · Suva · See the Fiji profile · See the Suva profile
Evidence: Observational / pre–post Top 35% 67/100 · Ask Evidence Copilot about this practice
Digital Earth Pacific, run by the Pacific Community (SPC), turns open satellite data into AI-processed disaster-risk, coastline and mangrove products for 20 Pacific governments; by 2025 it had deployed or scaled nine regional products and trained over 220 people.
The Pacific Community (SPC) built Digital Earth Pacific on Microsoft's Planetary Computer, using AI and cloud processing to turn petabytes of openly available satellite imagery into products usable for national planning. Early applications included assessing volcanic-eruption damage to crops and water supplies, and helping identify relocation sites for communities displaced by cyclones.
By its 6th Steering Committee meeting, held in Nadi, the programme brought together representatives of 20 Pacific countries and territories — including Fiji, Papua New Guinea, Samoa, Solomon Islands, Tonga, Vanuatu, Kiribati and Tuvalu — with partnerships spanning Geoscience Australia, the Group on Earth Observations, Digital Earth Africa, and two regional geospatial councils (PGRSC and PGSC). Products cover coastline change, flood patterns, mangroves, seagrass, vegetation height and near-shore water depth.
The programme reported nine regional data products deployed or scaled and more than 220 people reached through training. One concrete product, the Digital Earth Pacific Mangroves Extent and Density dataset, tracks mangrove change across Pacific Island Countries and Territories from 2016 to 2022 with planned annual updates, is openly licensed under Creative Commons Attribution 4.0, and is hosted publicly on AWS.
Public reporting to date documents outputs and training reach rather than independently verified outcomes — the 2026 steering committee explicitly flagged 'strengthened national uptake' and 'clearer governance pathways' as unresolved priorities, so real-world decision impact is not yet independently measured.
Reusable artefacts from this practice — as published by their sources.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Tuvalu
Tuvalu's Department of Lands and Survey built a machine-learning-assisted digital twin of Funafuti under the Pacific Community's Digital Earth Pacific …
Samoa
Samoa's national weather service now issues impact-based forecasts, not just hazard alerts, for cyclones, coastal flooding and storm surge, under …
Kenya
Researchers at the University of Sussex and Kenya's RCMRD built machine-learning models forecasting NDMA's official VCI3M drought indicator up to …
Kyrgyzstan
Kyrgyzstan's Ministry of Emergency Situations runs a national AI module that analyses satellite imagery to track ice-cover and breach risk …
Open full copilot Grounded in cited practices — always check the sources.