Operational forest structure monitoring using airborne laser scanning

The structure of forests influences the global biochemical cycles and can serve as an indicator to estimate the conservation potential for biodiversity and to determine forest stand resistance to disturbances. Airborne laser scanning (ALS) systems have been proven as suitable tools to provide horizontal as well as explicit vertical information due to the canopy penetration of the emitted signal. We developed robust methods based on multi-temporal ALS data to provide a more efficient monitoring of forest structure variables. The derived forest structure information includes: i) the individual tree delineation, ii) the detection and description of understory and ground cover, and iii) the detection of dead wood. We used full-waveform ALS data under foliated and defoliated conditions in dense, deciduous dominated forest stands in the Lagern (Switzerland) and the Uckermark (Germany) Based on the ALS point cloud with the traditional geometrical characteristics and the related full-waveform information, we applied the following methods: i) hierarchical, 3D-clustering and alpha shape derivation for the individual tree delineation, ii) grid based, vertical stratification for understory detection, and iii) combination of map algebra and vectorization for the dead wood analysis. The validation showed high accuracies for the derived forest structure information following the requirements of traditional forest inventories. We conclude that it is possible to detect and characterize the forest structure with robust methods based on full-waveform ALS data; however, the availability of foliated/ defoliated ALS data with a high point density is indispensable.