Statistical analysis of airborne LiDAR data for forest classification in the Strzelecki Ranges, Victoria, Australia

Abstract

Although remotely sensed data have been widely explored for forest applications, passive remote sensing techniques are limited in their ability to capture forest structural complexity, particularly in uneven-aged, mixed species forests with multiple canopy layers. Generally, these techniques are only able to provide information on horizontal (two-dimensional) forest extent. The vertical forest structure (or the interior of the canopy and understorey vegetation) cannot be mapped using these passive remote sensing techniques. Fortunately, it has been shown that active remote sensing techniques via airborne LiDAR (light detection and ranging) with capability of canopy penetration yields such high density sampling that detailed description of the forest structure in three-dimensions can be obtained. Accordingly, much interest is attached to exploring the application of this approach for identifying the distribution of designated vegetation communities. However, the suitability of LiDAR data for the classification of forests with complex structures, particularly for cool temperate rainforest and neighbouring uneven-aged mixed forests in a severely disturbed landscape has hitherto remained untested. This study applied airborne LiDAR data for the classification of cool temperate rainforest dominated by Myrtle Beech (Nothofagus cunninghamii) and adjacent forests including naturally regenerated Mountain Ash (Eucalyptus regnans), mixed forest consisting of overstorey Mountain Ash and understorey Myrtle Beech, Silver Wattle (Acacia dealbata), and hardwood plantation dominated by Shining Gum (Eucalyptus nitens) in the Strzelecki Ranges, Victoria, Australia. LiDAR data were extracted within each of the forest plots. Non- ground laser returns were used to generate forest height profiles for the analysis of the spatial distribution of vertical forest structure for the plots dominated by different forest types. The k-means clustering algorithm was performed on each of the plots to stratify the vertical forest structure into three layers, representing the overstorey, mid-storey and lower storey of the plot-level forests. Variables were then calculated from the LiDAR data based on the three-layered structure for each plot. The statistical analyses, which included one- way ANOVA (analysis of variance) and the post hoc tests, identified effective variables for forest type classifications. Linear discriminant analysis with cross-validation was carried out to classify the forest types and assess the classification accuracy using error matrixes. This study demonstrated the applicability of airborne LiDAR for the classification of the Australian cool temperate rainforest and adjacent forests in the study area.