Calibration of GEDI L4A AGBD products in boreal and tropical forests of China

Accurate forest aboveground biomass estimation is essential for carbon accounting and ecosystem monitoring. NASA’s Global Ecosystem Dynamics Investigation (GEDI) provides a unique opportunity for large-scale estimation of aboveground biomass density (AGBD). However, the GEDI L4A products exhibit considerable uncertainty in China due to limited local training samples, and regional validation and calibration remain scarce. This study validates and calibrates L4A estimates using AGBD derived from unmanned aerial vehicle (UAV) observations and airborne laser scanning (ALS) data across two representative forest ecosystems, including the boreal forests of Daxing′anling in northeastern China and the tropical rainforests of Jianfengling in southern China. Results show that L4A AGBD is overestimated in Daxing′anling, with a bias of 22.33 Mg/ha, and underestimated in Jianfengling, with a bias of −87.62 Mg/ha. Intercept-based calibration reduces the root-mean-square error (RMSE) from 34.71 to 25.48 Mg/ha in Daxing′anling and from 96.54 to 38.01 Mg/ha in Jianfengling, while the calibration based on both intercept and slope yields slightly better performance, reducing RMSE further to 24.43 and 33.62 Mg/ha, respectively. These findings underscore the necessity of GEDI L4A model calibration and suggest that intercept-only calibration can effectively improve the accuracy of L4A products in some data-scarce regions.