Modelling aboveground forest biomass as a continuous distribution using terrestrial and aerial laser scanning: a pathway to enhance the precision of forest biomass estimates

Estimating forest biomass and carbon storage is a core goal of many forest monitoring programs. Biomass assessments typically rely on allometric models to predict individual tree biomass from sample plots. Commonly, predicted tree biomass is then assigned exclusively to the stem position and fully added to a plot if the tree base lies within plot boundaries. However, tree biomass–particularly crown biomass–is spatially distributed across the crown projection areas. Consequently, crowns of trees inside a plot may extend beyond plot boundaries, while crowns of trees outside the plot may partially overlap the plot area. These boundary effects are not considered in conventional approaches during plot-level biomass aggregates–leading to errors when field-based biomass estimates are co-registered with remotely-sensed metrics, which provide continuous spatial signals at defined resolutions over tree canopies. The resulting scale-mismatch can compromise model qualities and increase uncertainty in remote sensing–based biomass predictions. Here, we further developed the ‘horizontal biomass distribution ( HBD )’ approach–a solution that looks at the tree biomass strictly within the confinements of the plot boundaries by spatially explicit modelling biomass as continuous horizontal distributions. We developed and evaluated two groups of HBD models using crown and point cloud metrics: (1) considering simplifying theoretical assumptions: perfectly upright stem, a circular crown projection area, and an isotropic biomass distribution around the stem axis; and (2) models accounting for stem irregularities, crown shape-asymmetries, and spatial biomass variabilities. The models were derived from point clouds generated by fusing terrestrial laser scanning (TLS) and unmanned aerial vehicle laser scanning (ULS) datasets. Our results show that integrating crown and point cloud metrics yields a more realistic representation of crown biomass distribution by capturing structural asymmetries and biomass variability. Point clouds provide a practical means for describing HBD , which would otherwise be impossible through direct measurements. The HBD is expected to improve biomass estimates by enabling a geometrically consistent linkage between plot-level biomass estimates and remotely-sensed metrics.