evidoria

← Back to browse

Good practice Imported

Digital Earth Pacific — AI Satellite Monitoring for Coastline, Mangrove and Water Change Across the Pacific

New Caledonia · Nouméa · See the New Caledonia profile

Evidence: Descriptive / self-reported Top 25% 73/100 · Ask Evidence Copilot about this practice

The Pacific Community's Digital Earth Pacific, built on Microsoft's Planetary Computer, applies AI to satellite imagery to track coastline erosion, mangrove loss and surface water for every Pacific island member state and territory, free and open, since its October 2023 launch.

more than 150 million km²
Area covered by Digital Earth Pacific products
roughly 12.7 million people
Population covered across Pacific member countries and territories
27
Pacific member countries and territories covered
1.07 USD billion
Average annual disaster losses across Pacific island nations
449.4 USD million
Cyclone Pam damage (Vanuatu, 2015) (2015)
600 USD million
Cyclone Winston damage (Fiji, 2016) (2016)
Digital Earth Pacific — AI Satellite Monitoring for Coastline, Mangrove and Water Change Across the Pacific

Details

Maturity
Scaling
Promoter
Pacific Community (SPC), with Microsoft Planetary Computer, NOAA, the governments of New Zealand and the United Kingdom, and the Patrick J. McGovern Foundation
Period
2022–2024
Keywords
earth observation, satellite imagery, climate resilience, disaster risk, coastal monitoring

Context

The Pacific Community (SPC), the region's main intergovernmental scientific and technical organisation, launched Digital Earth Pacific in October 2023 at its 53rd Committee of Representatives of Governments and Administrations meeting in Nouméa, New Caledonia, applying AI and cloud computing built on Microsoft's Planetary Computer and Azure Machine Learning to freely available satellite imagery.

Objectives

The platform generates standardised environmental products, covering coastline change, mangrove extent and surface water resources, identically for every one of SPC's Pacific member countries and territories, with Director-General Stuart Minchin describing it as intended to become an operational public good that regularly generates products whenever a satellite flies over.

Activities

Funding and technical partners publicly credited include the governments of New Zealand and the United Kingdom, the US National Oceanic and Atmospheric Administration, the Patrick J. McGovern Foundation, and Microsoft, covering an area SPC describes as more than 150 million square kilometres and roughly 12.7 million people across 27 countries and territories.

Results

The initiative directly targets the region's disaster exposure: Pacific island nations face average annual disaster losses estimated at USD 1.07 billion, with single events such as Cyclone Pam (Vanuatu, 2015, USD 449.4 million) and Cyclone Winston (Fiji, 2016, USD 600 million) illustrating the stakes.

Conclusions

As of the most recent independent reporting in 2024, SPC officials described the platform as still working toward full operational status, and no third-party accuracy or outcome evaluation of the AI-generated products has yet been published.

Implementation

Indicative cost
Medium (€50k–€500k)
Time to results
Medium (1–3 years)
Staffing & skills
Pacific Community (SPC) — implementing intergovernmental body, Microsoft (Planetary Computer and Azure Machine Learning technical partner), Governments of New Zealand and the United Kingdom (funding partners), US National Oceanic and Atmospheric Administration (NOAA), Patrick J. McGovern Foundation (funding partner)

Conditions for success

  • Built on existing cloud and AI infrastructure (Microsoft Planetary Computer, Azure Machine Learning) rather than a bespoke system built from scratch.
  • Standardised product generation so every SPC member country and territory receives the same environmental products regardless of size or resources.
  • Multiple funding and technical partners spread cost and technical risk across the region.

Common failure modes

  • As of the most recent independent reporting in 2024, SPC officials described the platform as still working toward full operational status.
  • No third-party accuracy or outcome evaluation of the AI-generated products has yet been published.

Commonly funded by

Philanthropic / foundation funding

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful