Estimating aboveground biomass of low-stature Arctic shrubs with terrestrial LiDAR

Abstract:

Arctic tundra ecosystems appear to be responding to rapid climatic warming via changes in vegetation composition and increased woody biomass, which may induce significant shifts in ecosystem structure and function. Although understanding these shifts is important for predicting ecosystem trajectories, establishing methods for quantifying and scaling woody plant biomass in low-stature biomes is challenging. We used LiDAR data from a terrestrial laser scanner (TLS) to estimate harvested biomass and leaf area of two dominant low-stature (<1.5 m) Arctic shrub species (Salix pulchra Cham. and Betula nana L.) in small (0.64 m²) plots. We explored two biomass estimation approaches (volumetric surface differencing and voxel counting) applied to point clouds obtained from either close-range (2 m) or variable-range (<50 m) TLS scans. Relationships between harvested biomass and laser scan metrics were strong for all combinations of approaches. Voxel counting provided a marginally better result than surface differencing for close-range scans (R² = 0.94 vs 0.92; RMSE = 102 g vs 117 g), while surface differencing proved stronger than voxel counting for variable-range scans (R² = 0.90 vs 0.79; RMSE = 128 g vs 184 g). Strong relationships between total harvested biomass and total leaf dry mass (R² = 0.93; RMSE = 13.4 g), and between leaf dry mass and leaf wet area (R² = 0.99; RMSE = 9.01 cm²) justify estimation of shrub leaf area from TLS-derived shrub biomass. Our results show that rapidly acquired, repeatable terrestrial laser scans taken from multiple distance ranges can be processed using simple algorithms to yield aboveground biomass and leaf area estimates for low-stature shrubs at fine spatial scales (sub-meter to 50 + meters). These data have the fidelity required to monitor small but ecologically meaningful changes in tundra structure, and could be employed as ground reference data for broader scale remote sensing data collection to provide shrub biomass and leaf-area estimates at fine resolution over large spatial extents.