Advanced remote sensing technology for forest inventory
Gaia Vaglio Laurin
DIBAF department, University of Tuscia, Viterbo, Italy
Recent advancements in remote sensing technology opened new roads for forest inventory and parameters prediction, allowing to extrapolate local field measures to large areas with increased accuracy. The relevance of forest parameters, such as biomass and biodiversity, to the global carbon cycle is well known. Here the functioning and role of lidar sensor (light detection and ranging) is reviewed to illustrate how this tool, in conjunction with ground records and satellite imagery, can be employed to map above ground biomass over large regions. Specifically, lidar point cloud provides different height metrics that are used to establish robust regressions with stem biomass. Lidar airborne acquisitions can be realized as transects over pre-stratified forested regions; these lidar data strips, and the derived biomass predictions, calibrated with limited field plots, can then be used as surrogate ground truth for predictions over entire regions using as upscaling support additional optical and SAR satellite data. Optical data at current spectral and radiometric resolution offer the opportunity to discriminate among various forest types also detecting disturbance at fine scale; SAR advanced polarimetric information can be directly related to forest volume. The SAR and optical merged information, together with the abundant surrogate ground truth provided by lidar, are the basis to build accurate biomass prediction at country and regional levels. Lidar has been used also to predict biodiversity, as the vertical structure of the forest scanned by lidar in 3D is in direct relationship with the number of species of certain taxa (birds, insects), whose presence is influenced by the vegetation structure. In addition, lidar can support optical data to understand forest disturbance and canopy species biodiversity.