Development of lumber recovery correction models for plantation-grown Pinus banksiana trees

Abstract

The prediction of product recovery based on tree size usually does not take stem deformations into account, potentially leading to an overprediction of product recovery. Using first order, pure and full second and third order polynomial functions, this study quantified the difference in lumber volume recovery between real stems processed in a real stud mill and virtual stems estimated using a sawing simulator in relation to diameter at breast height (dbh, D) and total tree height (H) for plantation-grown Pinus banksiana Lamb. The actual lumber recovery was only 64% (volume) or 58% (value) as much as the virtual recovery due to stem deformations. The overestimation of lumber recovery using virtual trees is highly and positively correlated with dbh and tree height, and can be well estimated using D 2 or D 3 when only dbh is available or D 2 H when both dbh and tree height are available. Prediction error increased with increasing dbh. Through a simple transformation, the models serve as correction models to rectify the overestimation of virtual lumber recovery using forest inventory data.