Individual Tree Crown Delineation and Species Classification in Urban Landscapes from Multi-Source Remote Sensing by Integrating SAM and Watershed Segmentation

To quantify forest carbon stock and its future potential accurately, we generated a 30 m resolution forest age map for China in 2020 using multisource remote sensing datasets based on machine learning and time series analysis approaches. Validation with independent field samples indicated that the mapped forest age had an R² of 0.51--0.63. Nationally, the average forest age is 56.1 years (standard deviation of 32.7 years).