Small–footprint LiDAR Estimation of Structural Properties of Woody Plant Communities in Complex Terrain, North-eastern Australia

Abstract

LiDAR remote sensing can be considered a key instrument for studies related to quantifyingthe vegetation structure. We utilised LiDAR metrics to estimate plot-scale structuralparameters of subtropical rainforest and eucalyptus dominated open forest in topographicallydissected landscape, in North-eastern Australia. This study is considered an extremeapplication of LiDAR technology for structurally complex subtropical forests in complexterrain. Thirty-one LiDAR metrics of vegetation functional parameters were examined.Multiple linear regression models were able to explain 62% of the variability associated withbasal area, 66% for mean dbh, 61% for dominant height and 60% for foliage projective coverin subtropical rainforest. In contrast, mean height (adjusted R2 = 0.90) and dominant height(adjusted R2 = 0.81) were predicted with highest accuracy in the eucalyptus dominated opencanopy forest. Nevertheless, the magnitude of error for predicting structural parameters ofvegetation was much higher in subtropical rainforest than those documented in the literature.Our findings reinforced that obtaining accurate LiDAR estimates of vegetation structure is afunction of the complexity of horizontal and vertical structural diversity of vegetation.