Improving the Accuracy of Aboveground Biomass Estimation of Natural Secondary Forests Using Individual Tree Features

Aboveground biomass (AGB) is vital for assessing productivity and carbon storage. The area-based approach (ABA) and individual tree approach (ITA) are commonly used in remote sensing for AGB estimation, but their combined application remains underexplored. This study introduced a practical method that combined ABA and ITA using airborne laser scanning (ALS) data to estimate AGB of natural secondary forests of northeastern China. The results demonstrated that combining ABA and ITA features significantly enhanced AGB estimation accuracy. CNN achieved the highest accuracy and substantially improved compared to using ABA features alone. Among ITA features, tree height-related features were the most influential, and combining multiple types of ITA features proved to be more successful than using a single type. Therefore, using CNN based on ABA and all selected ITA features is recommended in practical applications. This method simplifies the complex ITA process and can be applied for large-scale, efficient forest inventory, sustainable forest management, and precise forest monitoring and protection in the future.