It is critical to commercial forestry to know the volume of wood that can be extracted from a forest. This is approached by collecting inventory data about the trees in the field which includes, diameter of trees at breast height (DBH), tree height, the number of trees per unit area, stem quality and pruned height. These metrics are either directly or indirectly related to the wood volume. Over the past 3 decades, Airborne Laser Scanning (ALS) data have been established as an efficient tool for accurate forest inventory. Since the ALS incurs high operational and logistic costs, there is a need to explore alternative resources. Advances in computer software and digital cameras have resuscitated the traditional Aerial Photography (AP) by being able to extract 3D point cloud data from high-overlapping digital aerial photographs. Estimates of forest inventory metrics from AP and ALS point cloud data have been reported to be of comparable accuracy. The objectives of this study are:
- To characterise and compare ALS and AP point clouds acquired over Pinus radiata plantations
- To compare individual tree and plot level forest metrics measured from ALS and AP point clouds
- To compare the inventory metrics derived using tree and plot level forest metrics from ALS and AP in order to assess the suitability of AP for forest inventory
This study also looks into the different workflows used to generate point cloud data from digital aerial photographs and compare the outputs with the benchmark ALS.
Figure 1 Orthophoto of a field plot (PHI1) with the plot boundary (orange circle) and a 1 m wide transect (yellow) running from West to East. The figure in the centre shows a side-view of an overlay of the different point clouds. A profile view along the transect is shown in the right figure
Figure 2 Canopy Height Models (CHMs) generated from the point clouds. From left to right, the first three CHMs were generated from AP point cloud data; whereas, the rightmost CHM is a product of the ALS point cloud.