Abstract:

Forests constitute the largest carbon reservoir in terrestrial ecosystems, and accurately assessing their carbon sink capacity is essential for maintaining the global carbon balance. However, forest ecosystem carbon sinks exhibit pronounced spatiotemporal heterogeneity, making their accurate quantification highly challenging. By integrating multisource information technologies, smart forestry provides new opportunities to overcome the limitations of conventional static monitoring and to support dynamic simulation and precision management. This study systematically reviews the major methods used for forest carbon sink estimation, summarizes their development trends and research hotspots, and provides a comprehensive review of the opportunities and challenges associated with the development of smart forestry. The study results showed that the number of publications on forest carbon sink estimation has increased exponentially worldwide, forming a research landscape centered on China and the United States, with Western Europe and East Asia as major supporting regions. Forest carbon sink estimation has evolved from early reliance on field surveys and empirical statistical methods toward the integrated application of process-based models, remote sensing-driven models, ecosystem service models, and artificial intelligence (AI)-assisted approaches. There are clear differences in research priorities are observed between international and Chinese studies: international research focuses more on climate-response mechanisms at the global scale, while research in China places greater emphasis on high-resolution regional mapping in support of the national carbon peaking and carbon neutrality goals. Forest carbon sink assessment has entered a new stage characterized by multi-technology integration and cross-scale coordination. Future research should focus on reducing uncertainties associated with scale transformation and improving model interpretability by applying edge computing to dynamic carbon sink monitoring and risk early warning, strengthening the coupling of ecological process mechanisms with deep learning, and promoting the coordinated use of satellite, airborne, and ground-based observations within digital twin platforms. These advances will help improve the timeliness, accuracy, and decision-support capacity of forest carbon sink estimation.

Key words: smart forestry, forests carbon sinks, digital twin, multi-source remote sensing fusion, machine learning