Abstract
Leaf Area Index (LAI) is a significant indicator of the forest dynamics and the ecological processes such as, the balance of global carbon exchange, the energy cycle in photosynthesis, evapotranspiration mechanisms, and water/nutrient cycling. LAI can be calculated based on the 3D point cloud data (PCD) collected by Terrestrial Light Detection and Ranging (LiDAR). The methods presented in literature calculate LAI indirectly and by applying the atudliary models and parameters such as radiative transfer modelgap fraction or the extinction coefficient. Conventionally, the LiDAR PCD is confined to a large cubic box and divided to a 3D array of voxels with uniform and arbitrary size. This generates a large number of empty voxels without any cloud points (OFF voxels). These OFF voxels are counted to calculate the gap fraction with a large error threshold. This paper addresses the mentioned drawbacks and presents a new simple and direct method for LAI calculation. In proposed method, intermediate auxiliary models and parameters are avoided. The LiDAR PCD is voxelized adaptively based on the variable local spatial point density and also the natural characteristics of the canopy. It does not require the large confining box around the tree PCD and counting of OFF voxels. In addition, the effect of varying laser beam diameter is considered in calculations. This method is applied to a Canadian boreal forest and the sensitivity of the calculated LAI to different parameters such as voxel size is analyzed.
NSTL
Elsevier