Estimation of Broadleaf Tree Canopy Height of Wolong Nature Reserve Based on InSAR and Machine Learning Methods

A novel approach combining Interferometric Synthetic Aperture Radar (InSAR) with machine learning demonstrates high accuracy in estimating broadleaf tree canopy height, a critical parameter for assessing forest carbon sinks. The study achieved root-mean-square errors of 1.76 m and 1.86 m using Random Forest, outperforming the Multi-layer perceptron method with errors of 3.96 m and 4.13 m. This cost-effective technique leverages UAV photogrammetry for true canopy height validation and integrates topographic factors to enhance model stability. The findings highlight the potential of InSAR and machine learning for large-scale, efficient forest monitoring, offering a robust tool for carbon cycle research in Wolong Nature Reserve.

Liu, X., He, L., He, Z., & Wei, Y. (2022). Estimation of Broadleaf Tree Canopy Height of Wolong Nature Reserve Based on InSAR and Machine Learning Methods. Forests, 13(8), 1282. https://doi.org/10.3390/f13081282