Anais do XIX Simpósio Brasileiro de Sensoriamento Remoto

Forest plantations cover a large part of tropical countries such as Brazil and eucalyptus plantations in particular account for 57% of Brazil’s reforested area. As plantations offer multiple benefits – such as an option to offset natural forests, simplify otherwise complex forest ecosystems, meet energy, pulp and paper demands, restore ecological services and combat climate change by sequestering carbon – to the society, monitoring and tracking the growth and productivity of forest plantations should be given high priority. In this regard, remote sensing techniques have been found highly efficient. The core objective of this study is to estimate individual tree attributes, such as tree height, diameter at breast height (dbh) and above ground carbon (AGC) stocks of eucalyptus plantations from lidar data using linear mixed effects (LME) models; Ordinary Least Square (OLS) regression models are also built for comparison purposes. From our results, it can be inferred that hierarchy existing within the plantation datasets can be well handled by LME models and predictive models, for tracking tree level AGC and forest productivity, with satisfactory accuracies possible by combining lidar and LME modeling techniques.