Convened by GEO, in partnership with the Maldives Ministry of Climate Change, Environment and Energy (MoCCEE), in collaboration with Allen Institute for AI (AI2), United Stated Geological Survey (USGS), Planet Labs, and Esri.

The Global Ecosystems Atlas Maldives Accelerator is a collaboration between the Group on Earth Observations (GEO) and the Ministry of Climate Change, Environment, and Energy (MoCCEE) of the Maldives. 

Supported by Ai2, USGS, Planet Labs, and Esri, the initiative leverages cutting-edge satellite imagery and AI to develop the first-ever comprehensive map of ecosystems in the country. 

The goal is to improve spatial planning, conservation efforts and biodiversity monitoring, covering marine, terrestrial and freshwater systems. 

“We need to ensure that we use our natural resources sustainably, protect and effectively manage ecologically significant areas, and restore degraded areas. We welcome the Global Ecosystems Atlas."

Building on existing work implemented by the International Union for Conservation of Nature (IUCN) in 2019, a kickoff workshop brought together national stakeholders and international experts to explore approaches to data discovery, ecosystem modelling and map production by two independent mapping teams, USGS and Ai2.

 During the workshop, Mark Otterlee from Ai2 demonstrated the critical role image annotation plays in training the institute’s AI models to classify ecosystem types.  

Workshop participants from 11 different ministries, agencies and universities and overseas interns and staff from the Global Ecology Lab, James Cook University, contributed to providing training data to this accelerator.

By combining high-resolution satellite imagery with advanced AI models, Ai2’s process identifies and classifies diverse ecosystem types across the Maldives by integrating multiple data sources, including field observations, to create a detailed and reliable map of the Maldives’ ecosystems. 

This will lead to the creation of a scalable, AI-powered mapping tool that not only provides detailed insights into ecosystems but can also be applied to Small Island Developing States and other regions, helping stakeholders make better decisions faster. 

The Maldives Accelerator has already yielded valuable insights, and has generated a first iteration map of terrestrial and shallow water ecosystems in the country. This has been a community effort, with a volunteer based Maldivian group having a vital part to play in helping Ai2 to train their foundational AI model.   

Upcoming activities will focus on further advancing the accuracy and validation of these maps, with an emphasis on field data integration to validate and refine the maps. 

Next steps are to refine the ecosystem models, expand data collection and integrate the map into national policy frameworks over the next 18 months. 

The plan includes generating a national Red List of Ecosystems and supporting national biodiversity targets. The data will also be accessible for disaster risk reduction, sustainable fisheries management and protected area planning.  

The Global Ecosystems Atlas Maldives Accelerator is a model that can be replicated to other countries facing similar ecosystem management challenges. In 2025, we continue to work with select countries to support in the development of ecosystem maps where they do not yet exist.