Critical analyses when modeling tree biomass to ensure additivity of its components.

It is presented the theme additivity of biomass of tree components. To evaluate and discuss this context, experimental information collected in forests of Acacia mearnsii De Wild. was used. Equations for components (stem and crown) and total biomass were fitted by means of two procedures: 1) generalized nonlinear least squares and 2) weighted-nonlinear seemingly unrelated regressions. Analyzing the performance of the estimators, it can be concluded that the two tested procedures are equivalent. On the other hand, this conclusion differs when evaluated the consistency and efficiency of the estimators. Fitting equations for the components and for the total biomass by an independent way is not realistic, because from a biological point of view the estimates of biomass are inconsistent, i.e., are not additive. The biomass estimates of the components and of the total, resulting from equations adjusted by means of systems of equations, provided narrower confidence intervals in relation to the equations adjusted independently, and is therefore more efficient. The second procedure presents better biological properties and statistics to estimate allometric equations for biomass of the components and for the total when compared with the independent estimation, thus it should be the method to be used.

Comparison of data mining and allometric model in estimation of tree biomass.

BACKGROUND: The traditional method used to estimate tree biomass is allometry. In this method, models are tested and equations fitted by regression usually applying ordinary least squares, though other analogous methods are also used for this purpose. Due to the nature of tree biomass data, the assumptions of regression are not always accomplished, bringing uncertainties to the inferences. This article demonstrates that the Data Mining (DM) technique can be used as an alternative to traditional regression approach to estimate tree biomass in the Atlantic Forest, providing better results than allometry, and demonstrating simplicity, versatility and flexibility to apply to a wide range of conditions. RESULTS: Various DM approaches were examined regarding distance, number of neighbors and weighting, by using 180 trees coming from environmental restoration plantations in the Atlantic Forest biome. The best results were attained using the Chebishev distance, 1/d weighting and 5 neighbors. Increasing number of neighbors did not improve estimates. We also analyze the effect of the size of data set and number of variables in the results. The complete data set and the maximum number of predicting variables provided the best fitting. We compare DM to Schumacher-Hall model and the results showed a gain of up to 16.5% in reduction of the standard error of estimate. CONCLUSION: It was concluded that Data Mining can provide accurate estimates of tree biomass and can be successfully used for this purpose in environmental restoration plantations in the Atlantic Forest. This technique provides lower standard error of estimate than the Schumacher-Hall model and has the advantage of not requiring some statistical assumptions as do the regression models. Flexibility, versatility and simplicity are attributes of DM that corroborates its great potential for similar applications.