Forest inventories are the foundation for forest management and the monitoring of forest ecosystems. Laser scanning from different platforms is in this context an important technology that presents a potential to supplement inventory data. In this work we connected and contrasted data from classical inventories with aerial and mobile laser scanning (ALS and MLS) for 11 sampling plots, with a focus on comparing different methods for growing stock estimation. We compared field measurement durations, detection accuracy, diameters at breast height (DBH), horizontal distances from plot centre, individual tree volumes and plot growing stocks. MLS proved statistically significantly faster compared to classical inventories, with an average duration of 15.4 minutes compared to 39 minutes. With MLS we detected 87.6 % of all trees, with lower detection for smaller trees. The RMSE in DBH values was 1.8 cm (6.6 %), with no statistically significant differences. DBH were overestimated for smaller and underestimated for larger trees. The RMSE in horizontal distances was 14 cm (1.6 %), with statistically significant differences. The impact of calculation method for tree volume estimations was larger than the DBH input source, with both variables being statistically significant. The impact of calculation method was the only statistically significant variable impacting growing stock estimations.
|
