Mapping previously undetected trees reveals overlooked changes in pan-tropical tree cover
Shidong Liu, Jie Zhang, Li Wang, Philippe Ciais, Jianjun Zhang, Josep Peñuelas, Biswajit Nath, Iskid Jacquet, Xia Wu, Shengping Ding, Li Wang, Ni Huang, Wanjuan Song, Wenjian Ni, Zheng Niu, Wenjian Ni, Zheng Niu
Abstract
Detecting tree cover is crucial for sustainable land management and climate mitigation. Here we develop an automatic detection algorithm using high-resolution satellite data (<5 m) to map pan-tropical tree cover (2015-2022), enabling identification and change analysis for previously undetected tree cover (PUTC). Our findings reveal that neglecting PUTC represents 17.31 ± 1.78% of the total pan-tropical tree cover. Tree cover net decreased by 61.05 ± 2.36 Mha in both forested areas (63.93%) and non-forested areas (36.07%) between 2015 and 2022. Intense changes in tree cover are primarily observed in regions with PUTC, where the World Cover dataset with a resolution of 10 m often fails to accurately detect tree cover. We also conduct a sensitivity analysis to quantify the contributions f climate factors and anthropogenic impacts (including human activities and land use cover change) to tree cover dynamics. Our findings indicate that 43.98% of tree cover gain is linked to increased precipitation, while 56.03% of tree cover loss is associated with anthropogenic impacts. These findings highlight the need to include undetected tree cover in strategies combating degradation, climate change, and promoting sustainability. Fine-scale mapping can improve biogeochemical cycles modeling and vegetation-climate interactions, improving global change understanding.