Speaker: Leonie Hodel, Research Fellow, University of Cambridge Geography Department
Deforestation in the Brazilian Amazon continues at alarming rates, with large areas of forest still being converted into pasture for cattle ranching. Despite a wide range of policy interventions, including the Brazilian Forest Code, intensification incentives, and supply-chain regulations, effective monitoring of cattle production remains a critical gap. Recent advances in artificial intelligence (AI) and very high-resolution satellite imagery open new possibilities to directly observe cattle at scale.
In this talk, Hodel presents an AI-based approach to detect and count cattle from satellite images and explore what these data can reveal about land-use dynamics at the deforestation frontier. She will demonstrate how cattle-level monitoring can be used to assess intensification patterns, detect grazing incursions into protected areas and Indigenous lands, and identify herds that fall outside existing supply-chain monitoring systems.
Beyond technical performance, she discusses how such data can be meaningfully integrated into current policy frameworks to support enforcement and transparency, while shedding light on the structural barriers to a just transition toward more sustainable and inclusive land-use systems in the Amazon.