In this paper it is shown that a simultaneous adjustment provides more efficient estimates of total tree biomass than with independent modelling for biomass estimates by compartments (canopy, bole and roots).
When modeling tree biomass, it is important to consider the additivity property, since the total tree biomass must be equal to the sum of the biomass of the components.
The aim of this study was to assess the simultaneous estimation performance, considering the additivity principle with respect to independent estimate when modeling biomass components and total biomass.
Individual modeling of total biomass and biomass components of leaves, branches, bole without bark, bole bark, and roots was performed on Pinus elliottii Engelm trees derived from forest stands in southern Brazil. Five nonlinear models were tested, and the best performance for estimating the total biomass of each component was selected, characterizing the independent estimation. The models selected for each component were fitted using the nonlinear seemingly unrelated regression method, which characterizes simultaneous estimation.
Independent fitting of coefficients for biomass components and total biomass was not satisfactory, as the sum of the biomass component estimates diverged from the total biomass. This was not observed when the simultaneous fitting was used, which takes into account the additivity principle, and resulted in more effective estimators.
The simultaneous estimation method must be used in modeling tree biomass.