Abstract: The population classification and sample structure on modeling of single-tree biomass equations for national biomass estimation in China was studied in this paper. Firstly, combining the classification of eco-geographic region and administrative region, a scheme of population classification on modeling of single-tree biomass equations was presented, which is mostly compatible with the population classification on modeling of single-tree volume equations, including 34 tree species or species groups, 6 large-scale geographic regions, and a total of 70 populations. Secondly, based on the analysis of variation coefficients derived from one-and two-variable tree biomass models fitted with available tree biomass data of several tree species, it was concluded that prediction precision of tree biomass models could be defined as more than 95%, and at least 150 sample trees are needed for a regional model which should be selected in equal numbers for each diameter class, i.e., not less than 15 trees in each of 10 diameter classes. Finally, by using the 7th National Forest Inventory data, the sample structure of each population was determined, that is, all the sample tree numbers of 70 populations were specified by province and by diameter class. The results will provide a practical reference for the establishment of generalized single-tree biomass models in China.

Key words: single-tree biomass equation modeling, population classification, sample structure, eco-geographic region, coefficient of variation, prediction precision