Land & Carbon Lab

🛰️ We are entering a new era in planetary data and intelligence. Join us for AI, Meet Nature — a special event introducing our powerful new AI tool designed to make environmental insight more accessible, intuitive and actionable — regardless of technical expertise. 📡 Register to join the livestream and be a part of this moment: https://lnkd.in/gbZSjrxy 📅 September 17 | 3:00–4:30 PM ET Co-hosted by Land & Carbon Lab, Bezos Earth Fund and World Resources Institute

📢 Another big data release from Land & Carbon Lab and the Global Pasture Watch consortium — that’s two in two weeks. 🌾Grasslands, savannas, shrublands and tundra cover nearly half of Earth’s land surface. These open ecosystems play a critical role in regulating water and carbon cycles, supporting biodiversity and sustaining the world’s grazing lands. Yet until now, global vegetation height maps were mostly built for forests — capturing tall tree canopies while missing the structural complexity of shorter ecosystems (1-7 meters). Today we’re closing that gap with the first global annual maps of short vegetation height — at 30 m resolution, spanning 2000–2024. ➡️Read the blog & access the data: https://lnkd.in/dMzU43Jg By putting short vegetation on the map, this open dataset creates new opportunities for researchers, conservationists, policymakers and restoration planners to monitor change, strengthen carbon accounting and plan smarter land use worldwide. 🙏A major thanks to our partners who made this possible: Lapig - Laboratório de Sensoriamento Remoto e Geoprocessamento, OpenGeoHub Foundation and NASA - National Aeronautics and Space Administration #OpenData #Grasslands #EcosystemMonitoring #GlobalPastureWatch

Most AI vision models need months of training and labeling of data before they can work effectively. 🚀 DINOv3 is different. Developed by AI at Meta with input from World Resources Institute, it learns from a massive set of images in advance, so it’s ready to use immediately — and performs better than models built for specific tasks. For Land & Carbon Lab, that means applying it to satellite imagery to deliver faster, cheaper and more reliable restoration insights: spotting new tree growth just months after planting and tracking recovery over time at a fraction of the usual cost — with the same breakthrough soon strengthening other monitoring tools too. 🌱 With support from the Bezos Earth Fund, we’re applying this leap in AI vision to make land monitoring more reliable and accessible for restoration leaders worldwide. Learn more: https://lnkd.in/dT7kb8_b #OpenData #AI #Restoration

🌾 How healthy are the world’s ecosystems— and how are they changing? Using two decades of Landsat imagery and a light use efficiency (LUE) model, Land & Carbon Lab and the Global Pasture Watch research consortium have created the first global maps of Gross Primary Productivity (GPP) at 30 m resolution. ➡️ Explore applications and access the data set here: https://bit.ly/45sM8q6 This 30-meter resolution data set captures how ecosystems convert sunlight to carbon during photosynthesis offering unprecedented detail for monitoring carbon uptake, vegetation health and land productivity. What you can do with it: 📈 Track carbon uptake by plants over time. 🌱 Detect shifts in productivity to spot early signs of stress or land degradation. 🦋 Support conservation planning by monitoring productivity in natural and semi-natural grasslands 🌾 Strengthen sustainable agriculture by using these as an input to crop yield mapping. By revealing fine-scale processes across more than two decades (2000-2024), this data provides the most detailed view yet of ecosystem productivity worldwide — helping decision-makers and practitioners translate global science into local action. #EcosystemMonitoring #ClimateData #OpenData

🚨 New data release from Land & Carbon Lab and the Global Pasture Watch research consortium: For the first time, farmers, land managers, policymakers and scientists can track land productivity worldwide at 30-meter resolution — with a consistent bi-monthly record spanning more than two decades (2000–2024). ➡️ Explore the data, methodology and applications here: https://bit.ly/45sM8q6 This new Gross Primary Productivity (GPP) data set shows the most detailed global view yet of the rate at which plants take in carbon during photosynthesis. Validated with more than 500 flux towers, the data capture fine-scale variation in ecosystem productivity that coarser products have missed - especially in underrepresented landscapes like grasslands and rangelands. 🌱 Why it matters: Critical metric for understanding land carbon dynamics. Sharper insights into land health, restoration potential and responses to climate change. Assessment of seasonality and long-term trends in vegetation productivity. 🙏 A major thanks to the Global Pasture Watch research consortium and partners who made this possible: OpenGeoHub Foundation Lapig - Laboratório de Sensoriamento Remoto e Geoprocessamento CMCC Foundation - Centro Euro Mediterraneo sui Cambiamenti Climatici CIFOR-ICRAF Alliance of Bioversity International and CIAT #EcosystemMonitoring #ClimateData #OpenData

We’ve just added a new page to the AFi website on the topic of non-forest ecosystem conversion! ✨ Non-forest ecosystems (like grasslands and savannahs) are vital for biodiversity, climate resilience, and many livelihoods across the globe. They are subject to many of the same pressures as forests, under threat from degradation and agricultural expansion—but they are often overlooked in sustainability efforts. We’ve been working with World Resources Institute (WRI) for the past 18 months to lead the way on mapping and protecting these types of ecosystems.  The new web page:  🌿 Outlines the importance of non-forest ecosystems  📑 Shows how the Accountability Framework can support companies meet their no-conversion commitments  💡 Features useful resources, including outputs from our work with WRI Check out the new web page here 👇  https://lnkd.in/ev3emSrS

DINOv3, developed by AI at Meta with input from World Resources Institute, is a breakthrough in AI vision models — and it’s setting a new standard for environmental monitoring. With support from Bezos Earth Fund, this is how we use it to map the height of individual trees globally — faster and more accurately than ever before: 🛰️ Satellite imagery – We start with high-resolution, cloud-free images of the Earth’s surface from across the globe. 🤖 AI vision analysis – DINOv3 pre-trained on satellite imagery detects tree cover and pulls out patterns linked to tree structure and height. 📏 Height calculation – The model calculates the distance from the satellite to the ground and from the satellite to the top of the canopy, then finds the difference to determine tree height. 🌱 Restoration insights – We take those findings to show where trees are growing, how tall they are (even from a young age) and how restoration sites are performing over time. For restoration monitoring, this is a massive leap forward — delivering higher accuracy and reliability in detecting and measuring trees without the task-specific training and months of labeling older models required. These advances strengthen our ability to track growth, estimate carbon storage and guide restoration investment — helping partners track progress and prove impact with broader applications ahead for biodiversity, agriculture and more. Learn more: https://bit.ly/4oyZqZs

🌱 Restoration needs data that moves as fast as the work on the ground. That’s why AI at Meta’s new, open-source DINOv3 vision model for satellite imagery — paired with innovation at Land & Carbon Lab and on-the-ground restoration impact through World Resources Institute — is a breakthrough. It brings sharper, faster insights to track where trees are growing, how tall they are (even from a young age) and how degraded areas are recovering — showing restoration performance and where to focus resources for the biggest impact. This leap in AI vision delivers higher accuracy and reliability while seamlessly monitoring across imagery from satellites, drones and smartphones — and does it without the long, costly task-specific training older models required. That means clearer, faster insights for funders to invest with confidence and for local partners to act before restoration opportunities are lost. Supported by the Bezos Earth Fund, this work is driving change where it matters most. And restoration is only the beginning — the same breakthrough will soon strengthen monitoring for biodiversity, agriculture and more. Learn more: https://bit.ly/4oyZqZs