Thinning modelling caught between algorithm and tree-marking.

摘要

In the wide range of thinning models, only a certain type is applicable to the specific demands of diverse mixed stands. This study analyses the suitable empirical-statistical and deterministic single-tree thinning models. The error components are compared for the two models. Four exemplary thinning trials investigate the magnitude of the variance in tree-marking. Possible consequences and application of modelling in thinnings are given. It is shown that person-specific variation in tree-marking affects all types of thinning models. The higher the resolution of the validation scale (from aggregated stand variables to individual tree comparison), the greater the effect. This qualitative analysis is complemented by presenting four specific thinning models. Based on these examples, the requirements regarding simulation goals and the consequences from the variance in tree-marking are discussed..