Comparison of conventional measurement and LiDAR-based measurement for crown structures

Tree crown projection area and crown volume are the important parameters for the estimation of biomass and other forestry science applications. However, it is difficult to validate their accuracy. In this research, measurement errors were quantified based on the 3D virtual LiDAR dataset of 22 common tree species in Beijing, China, which were derived from high-resolution terrestrial laser scanner. The hypotheses include that: (1) primary sources of height measurement are the violations of Total station based on the triangulation theory; (2) approximation of tree crown projections with 4-point fixed angles are not precise enough for irregular crown silhouettes; and (3) tree crown volumes obtained by conventional method with a taper's formula (V"C) are not corresponding to the need of accurate measurement. The relative accuracies derived from total station measurements between crown base height (B"T) and tree height (H"T) were compared. Meanwhile, different methods for tree crown projections and crown volumes were researched. The results show that the measurement errors of H"T (RMSE=4.56m) is higher than that of B"T (RMSE=3.35m), V"C (RMSE=45.0243m^3) is 8.03% on average smaller than V"V"C (RMSE=48.56094m^3), A"4 is 25.5% on average larger than A"V. The deviations between filed investigation method and virtual 3D LiDAR data could be attributed to irregularity of the crowns' silhouettes. Based on the results, we propose that: (1) the use of 8-point or 16-point projection with fixed angles to estimate crown projections and (2) different regular volume formula to simulate crown volume according to the tree crown shapes.