Optimal sample trees diameter distribution in developing allometric biomass models

Abstract

Abstract Background: Improving forest biomass and carbon estimates is essential for sustaining the mitigation of climate change efforts in the forestry sector. An important source of uncertainty in forest estimates originates in the allometric model predictions. When developing allometric biomass models, the tree selection process is an important step that affects the model’s predictive performance. Typically, the diameter at breast height (D) range of the species is divided into D-classes, followed by random selection of sample trees within the D-classes; the number of trees in each D-class defines the sample tree D-distribution. Here, using a simulation study, we compared six types of sample tree D-distributions with respect to the precision of estimates of mean population biomass that the models produced. Results: The results showed that randomly selecting from each D-class a number of trees that is proportional to the basal area in that specific D-class in the population (i.e., sample trees were selected to produce a distribution of basal area in the sample that matched the distribution in the population) was the optimal D-distribution of the sample trees for minimizing the standard errors of the estimates of the population mean for a given sample size. When a-priori information about the distribution of tree attributes in the population is unknown, a uniform D-distribution represents a good alternative to the optimal sample D-distribution. Although producing the greatest precision of the estimate, the optimal D-distribution of sample trees did not produce models with the most accurate model fit (greatest coefficient of determination), suggesting that model fit alone is not a sufficient indicator of model’s predictive performance. Conclusions: The sample tree D-distribution affected considerably the precision of estimates of mean population biomass that the models produced. Therefore, to optimize the tree selection, and, therefore, to develop the models that produce precise estimators of the population mean, we recommend using an optimal D-distribution of the sample trees.