Validation of terrestrial laser scanning data using conventional forest inventory methods

Abstract

The application of terrestrial laser scanning (TLS) in capturing forest inventory parameters such as diameter at breast height, height and diameters along stem profiles, and in monitoring forest growth, was investigated and validated by comparison with conventionally measured individual tree parameters and plot-level forest growth in a stand of Sitka spruce (Picea sitchensis (Bong.) Carr.) in Ireland. The data acquisition for all the plots with different tree sizes and different slopes was carried out using a terrestrial laser scanner (FARO LS 800 HE80) in November 2007 and November 2009, using the same plot centres and measurement procedures. The point cloud data were processed with Autostem™ software. The results showed that TLS enables the acquisition of forest stand parameters with an acceptable accuracy. Pruning of the lower branches did not improve tree recognition and the number of (partly) occluded trees stayed the same. Over the 2-year period, the average difference between the volume increment of the trees visible to the scanner derived using the conventional method and Autostem™ was 4.77 m3 ha−1 and resulted in scanner-derived estimates that were lower than the estimates obtained by conventional method by 6.1 %. Using a simple correction factor to account for occlusion in the laser scanner data, the difference between these estimates for all trees in the stand became an over-estimation by 6.96 m3 ha−1 (8.1 %). At heights up along the stems >15 m, the errors in stem diameter estimates started to escalate.