A practical application of airborne LiDAR for forestry management in Scotland

This paper presents an assessment of the risk of windthrow for an area of ancient woodland of high environmental importance in Scotland, UK. ForestGALES (Geographical Analysis of the Losses and Effects of Storms in Forestry) is a process-based model, which identifies wind vulnerability. The collection of plot-level field data has previously permitted this model to be run for stand-level analysis. In this study, airborne LiDAR data were used to produce a normalised canopy height model (CHM). An algorithm was designed within Definiens Developer 7.0 object oriented analysis software in order to delineate individual tree crowns from the CHM. The results of this delineation were used to develop regression equations using individual tree height and crown width to estimate diameter at breast height. This aims to allow structural vegetation properties to be related to the spatial distribution of individuals. The spatial arrangement of individual tree heights and diameters at breast height were used to generate tree lists and use them as inputs to ForestGALES. This allowed the stability of individual trees to be mapped by modelling the critical wind speed at which they are predicted to be overturned. This offers a substantial improvement on previous model outputs and provides important data, which can inform forest management decisions.