• High-resolution tree data: The first AI-powered global map with 1-meter resolution provides detailed canopy height data, enabling detection of individual trees worldwide.
• Open access for impact: The data and models are free and publicly available, promoting innovation in carbon credit verification and conservation efforts.
• Transforming carbon monitoring: High-resolution mapping supports precise forest management and strengthens transparency in carbon markets and restoration projects.

Meta, the World Resources Institute (WRI), and Land & Carbon Lab have unveiled the first global AI-powered map of tree canopy height at 1-meter resolution. This groundbreaking tool enables detection of individual trees worldwide and addresses critical gaps in understanding forest ecosystems.

Revolutionizing Forest Monitoring:

The map leverages artificial intelligence to analyze over a trillion pixels from 18 million satellite images, creating a detailed global baseline of tree canopy height. According to Meta, the model’s mean absolute error is just 2.8 meters, making it highly accurate for monitoring and verification purposes.

“Democratizing access to artificial intelligence can be an important tool in unlocking finance for and increasing transparency in mitigating and adapting to climate change,” stated Meta.

The dataset reveals that one-third of Earth’s landmass, about 50 million square kilometers, has a canopy height above 1 meter. It is freely available on platforms such as AWS, Google Earth Engine, and GitHub for both commercial and conservation use.

Related Article: Multilateral Climate Funds Launch AI-Powered Climate Project Explorer at COP29

Supporting Carbon Markets:

The map addresses a key challenge in carbon removal strategies: accurate monitoring and verification of forest-based carbon credits. By enabling high-resolution tracking of tree growth, particularly in sparse or small-scale forests, it improves transparency and accountability in the carbon market.

“Forest-based carbon removal and the use of technology to better monitor, report, and verify carbon sequestration are essential components of Meta’s carbon removal strategy,” the company noted.

A Tool for Climate Action:

The AI model powering the map, known as DiNOv2, employs Self-Supervised Learning (SSL) to process unlabeled satellite images, ensuring robust global performance. This scalable technology also supports other applications, including tree detection and segmentation.

With this innovation, Meta and its partners aim to drive collaboration and actionable solutions for conservation, carbon market integrity, and global climate adaptation. This map represents a transformative step in harnessing AI for environmental resilience.

Follow ESG News on LinkedIn