Developed by WRI and Meta, the model provides unprecedented insight into trees outside of dense forests. The data is already being used to monitor small-scale restoration throughout Africa, which can help these projects access much-needed finance.
The Challenge
Forests play a vital role in combating climate change; providing water, food and medicines; and supporting livelihoods for millions of people. But monitoring them is challenging, particularly for trees that grow outside dense forests in drylands, farms and cities. These trees make up more than one-third of tree cover on Earth, yet they go largely undetected by forest data sets.
This data gap is especially limiting for monitoring restoration projects, much of which are carried out by individual farmers and local organizations revitalizing small plots of land. As a result, funders often prioritize larger, easier-to-monitor projects over the smaller, locally led projects proven to be more effective at restoring degraded landscapes. Overcoming this gap is essential to help local communities protect their environment and access restoration finance.
WRI’s Role
WRI’s Land & Carbon Lab initiative collaborated with Meta’s Fundamental AI Research and Sustainability teams to develop a first-of-its-kind AI-based global tree canopy map with 1-meter resolution. Meta provided advanced computational resources and AI modeling, while WRI contributed expertise in restoration monitoring and remote sensing.
WRI then facilitated the application of the data to restoration programs such as TerraFund, which funds 192 local restoration organizations across Africa. WRI also shared the model and its results through its extensive network, ensuring the data could benefit other organizations globally.
The Outcome
The ability to map every tree in the world at 1-meter resolution provides unprecedented visibility. It holds particular promise for monitoring restoration and tree-planting, making it feasible to monitor small-scale projects in rural, hard-to-reach areas that are often overlooked due to high monitoring costs. The AI-based approach is 10 times cheaper than traditional field-based monitoring.
The tree canopy height map already supports over 28,000 restoration sites across Africa through WRI’s TerraFund initiative, providing critical data to evaluate projects’ impacts on land restoration and carbon sequestration. Being able to see individual trees on the ground gives credit to the thousands of smallholder farmers revitalizing their land. This can convince investors that financing local restoration is an effective nature-based solution to climate change and rural poverty. The data also allows governments to better understand how these communities are contributing to national land restoration pledges and climate plans.
The data has also been integrated into global and national conservation efforts. Half a dozen countries are utilizing the dataset to set baselines for the EU Deforestation Regulation, while the EU Commission has integrated it into its updated global forest baseline map.