Reconstructions of Tree Structure from Laser-scans and Their Use to Predict Physiological Properties and Processes in Canopies

Two ways of tree structure reconstruction from laser-measurements are comparatively presented with respect to their suitability for functional interpretation of tree structures. A concept for physiological evaluation of the data is tested that comprises three steps: 1. The automated evaluation of terrestrial laser-scanner data, 2. application of a fine-scaled 3D-light model, and 3. a nitrogen dependent photosynthesis model. The automated evaluation procedure for laser-scanner data based on the 3D-Hough transformation could correctly identify the point-to-point connections of the main branch system of apple trees. Based on 3D-structure measurements, the 3D-light model STANDFLUX-SECTORS predicted maximum leaf mass per area (LMAmax) of branches with a root mean square error of 10.3 g/m2. Vcmax and Jmax of beech and oak leaves were shown to be derivable from LMA. The comparison between daily courses of modelled light and branch sapflow showed similarities, but also dissimilarities that can only be explained with fine-scaled 3D-structure. * Corresponding author.