Use of terrestrial laser scanning to update form factors for Dutch forests

By Dimitra Veropoulou

Abstract
Accurate tree volume estimation is essential for biomass, carbon stock calculations and therefore for forest inventories. The National Forest Inventory (NFI) of the Netherlands currently uses the Schelhaas function, the accuracy of which is not verified. There is a growing interest in utilizing terrestrial laser scanning (TLS) as an accurate, non-destructive alternative method to measure tree attributes in large-scale applications. This research evaluated the effectiveness of three TLS-automated modelling tools, Quantitative Structure Model (QSM), Forest Structural Complexity Tool (FSCT), and 3D Forest Inventory (3DFin), by comparing the estimated volumes with the values from the Schelhaas volume function for two species, Douglas Fir and Beech. The results indicated that QSM and 3DFIn provide trunk volumes that closely align with the Schelhaas function, especially for Beech trees. Opposite to FSCT which systematically underestimates the trunk volume, particularly for Douglas Fir, due to errors during the segmentation step. Key finding of this research was the variability of the models’ performance across the different study areas, which indicated the need to establish standard protocols for TLS data collection. The findings of this study highlighted the potential to incorporate the TLS and the automated tools into the NFI methodology, along with the Schelhaas function as an intermediate step. For future research, it was suggested to focus on training the FSCT model for specific tree species to see if the model can be improved. Furthermore, it should be investigated whether one volume function is suitable for the whole country or region-specific volume functions should be created. Finally, it is recommended to collect existing or new validation data with which all methods should be compared with and assess their accuracy.

Keywords: TLS; trunk volume estimation; QSM; FSCT; 3DFin; National Forest Inventory