Accuracy assessment of terrestrial laser scanning for estimating tree height, diameter at breast height, crown dimensions, and aboveground biomass
Abstract
Terrestrial laser scanning (TLS) provides dense three-dimensional observations of forest structure and has the potential to complement or replace labor-intensive conventional measurements for selected tree attributes. This study establishes an accuracy-assessment framework for estimating tree height, diameter at breast height (DBH), crown dimensions, and aboveground biomass (AGB) from TLS point clouds. A field design of 20–50 individually identifiable trees is proposed, with conventional measurements used as independent reference observations. TLS data are acquired from multiple scan positions, registered into a common coordinate system, filtered, ground-normalized, and segmented at the individual-tree level. DBH is extracted from horizontal stem slices at 1.30 m, tree height from normalized vertical extent, crown dimensions from crown-point geometry, and AGB from species-appropriate allometric equations using field- and TLS-derived predictors. Agreement is evaluated using the coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), mean bias error, relative error, and Bland–Altman limits of agreement. The framework also tests whether TLS-derived measurement errors propagate materially into biomass estimates. The study is designed to identify which structural variables can be measured reliably with TLS, where occlusion and crown complexity introduce uncertainty, and whether non-destructive point-cloud measurements provide sufficient accuracy for forest inventory, ecosystem monitoring, and conservation-oriented applications. To demonstrate the complete statistical workflow without misrepresenting unobserved field measurements, the numerical tables in this methodological manuscript are populated with an explicitly labelled simulated 36-tree demonstration dataset; these values must be replaced by measured observations before empirical submission.