Modelling Volume, Growth and Mortality

The purpose of this paper is to introduce a highly effi cient technique for deriving accurate and versatile stem curve models. The technique in question is based on relative stem diameters and relative stem heights and is subsequently called the relative stem curve method. The paper focuses on the following three main issues (i) the uniformity and variability of relative stem curves, (ii) the volume and commercial height estimation accuracy of relative stem curve models, and (iii) the number of sample trees required to derive relative stem curve models. It is shown, by an extensive set of stems measured in the fi eld and comprising 1964 stems, that average relative stem curves of different sized trees are highly similar and that the variability of relative stem curves is also quite modest. Consequently the volume and commercial and saw-wood height estimation accuracies are high and the required number of fi eld sample trees low. Standard errors varied from 10% to 7% for volume and the required number of sample trees from 1 to 15. It is proposed that a worldwide effort is launched in order to derive and utilize relative stem curve models for all commercially important tree species.