Spaceborne LiDAR is essential for accurate forest canopy height measurement, which is crucial for assessing forest biomass and carbon storage. Since its launch in 2018, the Global Ecosystem Dynamics Investigation (GEDI) has been focused on estimating forest canopy parameters. However, traditional decomposition methods for GEDI waveform data often face challenges in complex terrain, which can reduce accuracy. This study introduces a composite Gaussian model decomposition (CGMD) method to improve canopy height estimation of the selected forest farm. The results showed that: (1) Canopy height model (CHM) resolution had minimal impact compared to data quality and buffer settings; (2) The proposed CGMD method consistently outperformed traditional methods, with RMSE reductions of 8.2% to 33.8%; (3) A two-sample t-test confirmed a statistically significant improvement (p < 0.05); (4) Further application of this method to another study area for verification confirmed its robustness and generalizability. The CGMD consistently outperformed the traditional decomposition methods, with RMSE improvement ranging from 12.9% to 23.6%. These results indicate that the CGMD offers a promising approach for enhancing canopy height estimation accuracy, particularly in complex terrains.
Wang et al. (Mon,) studied this question.