Global patterns in forest carbon storage estimation: bibliometric analysis of technological evolution, accuracy gains and scaling challenges

Introduction Estimation of forest carbon (C) storage is essential for understanding the global C cycle, mitigating climate change, and developing carbon markets. However, systematic research on forest C storage estimation needs improving. Methods Herein, a bibliometric and content review of literature published between 2008 and 2025 was conducted to synthesize temporal and spatial trends and to identify methodological advances and gaps in forest C-storage estimation. Results The results revealed that environmental sciences accounted for the largest share of publications ( n = 718). The most productive institution and country were the Chinese Academy of Sciences ( n = 208) and the United States ( n = 691), respectively. Research progress in the field was categorized into three distinct stages since 2008. The early stage (2008–2012) was dominated by eddy covariance, satellite remote sensing, and airborne radar. The middle stage (2013–2017) was characterized by greater use of process-based and statistical simulation models. In the later stage (2018–2025), techniques such as random forest (RF), machine learning and biomass mapping became more widely used. Over this period, model performance improved substantially, especially the coefficient of determination (R 2 ) increased from 0.62 to 0.97 for the TRIPLEX-Flux C-exchange model and from 0.63 to 0.97 for RF models. Discussion Spatially, most studies addressed local-to-regional scales, whereas large-scale or global assessments remain limited. This synthesis clarifies methodological trajectories and persistent gaps that can guide the development and wider deployment of forest C-storage estimation approaches and support evidence-based climate policy and C-market design.