Capturing trends in forest structural complexity development using laser scanning techniques

Forest structural complexity reflects realized niche occupancy, capturing how effectively the vegetation utilizes available resources and provides habitats for species. This makes it a key indicator of forest ecosystem diversity, and an important characteristic to be monitored to facilitate sustainable forest management and conservation planning. Laser scanning has been recognized as a feasible technology for the characterization of heterogeneity in forest structure, reflecting its structural complexity. However, less is known about the capability of different laser scanning techniques to capture structural complexity development through time, and whether the cross-use of various data types and analysis methods yields consistent observations of the development. We aim to address this knowledge gap by investigating the capability of different laser scanning techniques to assess forest structural complexity development and evaluate whether comparable observations can be obtained regardless of the laser scanning technology used. The experiments were conducted across 49 sample plots within southern boreal forests in Evo, Finland. A 7–10-year monitoring period was captured using terrestrial laser scanning (TLS), and airborne laser scanning (ALS) at three different resolutions representing low (0.4-1 pts/m²), medium (15-28 pts/m²), and high (200-3600 pts/m²) point densities. Eight metrics were used for structural complexity characterization: mean canopy height, canopy rugosity, gap fraction, vegetation occupancy, vertical evenness (Shannon entropy), variability in crown area and tree height, and mean fractal dimensions (box-dimension) among trees. Comparison of observations of structural complexity development showed that gap fraction and Shannon entropy exhibited consistent development directions and similar metric change magnitudes across all the investigated laser scanning techniques. In contrast, metrics characterizing three-dimensional complexity, such as vegetation occupancy and mean box-dimension, were more sensitive to point cloud data characteristics. These findings provide insights into selecting appropriate laser scanning techniques and analysis methods to monitor forest structural complexity development for applications such as conservation planning.