Mapping the World's Forests with Greater Precision: Introducing Canopy Height Maps v2

Meta AI, in partnership with the World Resources Institute, has released Canopy Height Maps v2 (CHMv2), an open-source model for global forest mapping. The upgrade replaces the DINOv2 backbone with DINOv3, a self-supervised vision model pre-trained on SAT-493M, a large satellite imagery dataset. The model's R² accuracy improved dramatically from 0.53 to 0.86, with better detail and reduced bias for tall trees. The training dataset was expanded with geographically diverse lidar data, automated matching tools, and a specialized loss function. CHMv2 is already being used by UK Forest Research, the European Commission's Joint Research Centre, and US city planning initiatives for forest monitoring, carbon estimation, and urban cooling strategies.