Non-destructive Aboveground Biomass Estimation in Teak Plantations: Validating Terrestrial Laser Scanning and Quantitative Structure Models against Destructive Sampling

• TLS-derived QSM volumes showed no significant difference from destructive sampling (p = 0.911; RMSE < 0.016 m³). • The DBH² × H allometric model from QSM data achieved the highest accuracy (R² = 0.98, RMSE = 7.29 kg). • Generalized regional equations significantly deviated from site-specific models (RMSE = 31.79 kg). • TLS-QSM provides a reliable, non-destructive framework for carbon accounting in teak plantations. Aboveground biomass estimation is critical for sustainable forest management and carbon sequestration accounting. This study evaluates the performance of Terrestrial Laser Scanning combined with Quantitative Structure Models in teak ( Tectona grandis ) plantations in Lampang Province, Thailand. The research was conducted in two primary phases. The first part focused on validating the accuracy of the digital reconstruction process by comparing volumes derived from AdQSM and TreeQSM against actual measurements obtained from destructive sampling of 30 reference trees. Statistical analysis using the Kruskal-Wallis H test demonstrated no significant difference between the digital model volumes and the physical measurements (p = 0.911), with RMSE less than 0.016 m 3 . These findings confirm that TLS-QSM technology provides high precision and can effectively replace destructive sampling for tree volume measurement. The second part tested the efficiency of ten biomass estimation models. Equations developed from QSM data, utilizing different parameters and sample sizes, alongside a popular regional equation were compared against a reference equation derived from destructive sampling. These models were applied to a population of 980 trees to assess estimation consistency. The analysis revealed significant differences at the population level (chi-square = 60.6, p < 0.001). The localized QSM-based model (Y4), which incorporates combined parameters (DBH² × H), achieved the highest accuracy (R² = 0.98, RMSE = 7.29 kg). In contrast, the widely used regional equation (Y9) showed significant deviation with a substantially higher error (RMSE = 31.79 kg). In conclusion, this study demonstrates that TLS-QSM technology is not only a robust tool for non-destructive biomass estimation, but localized allometric equations developed from QSM data also offer significantly higher accuracy than generalized regional models. This approach provides a reliable framework for high-precision carbon accounting in teak plantations while eliminating the need for destructive harvesting.